Investigation of migration and spreading of cells
expressing modified talin1 forms
Master’s thesis
Taina Viheriälä
University of Tampere
BioMediTech
August 2014
ACKNOWLEDGEMENTS
This thesis was carried out in Vesa Hytönen’s research group, Protein Dynamics, in
BioMediTech at the University of Tampere. I would like to thank him for this great
opportunity to be part of a truly inspiring and interesting field of research. I would like
to thank my supervisor M.Sc. Rolle Rahikainen for guidance and for providing
solutions to problems during my everyday lab-work. I would also like to show my
gratitude to our laboratory technician Ulla Kiiskinen for her patience and exceptional
knowledge for almost anything I ever figured to ask.
I will also give my special thanks to Jenita Pärssinen who guided me whenever Rolle
was not around and to our other group members who helped me during my work with
specific and technical aspects; Teemu Ihalainen and Anssi Nurminen. But without the
whole group, the atmosphere would not be as great as it was so thank you all.
I am also grateful for my family and friends for their support during this thesis and their
encouragement throughout my study.
Finally, but most importantly, I want to thank my husband Tommi for giving me
exceptional support and showing that you always believe in me!
PRO GRADU-TUTKIELMA
Paikka:
Tampereen Yliopisto
BioMediTech
Tekijä:
Taina Viheriälä
Otsikko:
Migraation ja leviämisen tutkiminen soluilla, jotka ilmentävät muokattuja
taliini1 -proteiineja
Sivut:
61 sivua
Ohjaajat:
Dosentti Vesa Hytönen ja FM Rolle Rahikainen
Tarkastajat: Professori Markku Kulomaa ja dosentti Vesa Hytönen
Aika:
Elokuu 2014
TIIVISTELMÄ
Fokaaliadheesiot (FA) ovat suuria, useiden proteiinien muodostamia komplekseja. Solut
käyttävät adheesioita vuorovaikuttaakseen ympäristönsä sekä muiden solujen kanssa ja
niiden avulla solut pystyvät mm. liikkumaan. Yksi tärkeistä proteiineista, joka osallistuu
adheesioiden muodostamiseen ja ylläpitämiseen on taliini. Taliini muodostaa yhteyden
solun pinnalla olevan integriinin ja solun sisällä olevan aktiinitukirangan välille. Tämä
mahdollistaa solun vasteen ulkopuolella tapahtuvaan mekaaniseen muutokseen. Useat
FA-proteiinit on tunnettu jo pitkään, mutta niiden ominaisuuksista ja tehtävistä soluissa
on vielä puutteellisesti tietoa. Tässä työssä tutkittiin erilaisten lyhennettyjen taliini1 muotojen vaikutusta solujen migraatioon ja leviämiseen. Tätä ennen täytyi kuitenkin
optimoida ja pystyttää laitteisto migraatiokoetta varten. Optimointivaiheessa käytettiin
kolmea eri solulinjaa; villityypin hiiren alkion fibroblasteja (WT MEF) sekä soluja,
joista on poistettu taliini1:ä tai vinkuliinia koodaava geeni. Varsinaisessa migraatio- ja
leviämiskokeissa käytettiin hiiren alkion fibroblasti -soluja, jotka eivät ekspressoi
taliini1:ä, joten niihin pystyttiin transfektoimaan haluttua taliini1 -plasmidia; täyspitkää
taliin1:ä (FL Tal1) tai neljää erilaista lyhennettyä taliini1 –muotoa. Kaikki
taliiniptoteiinit ekspressoitiin fuusioituna punaiseen fluoresoivaan proteiiniin
(mCherry), joten fluoresenssimikroskoopilla oli mahdollista saada näkyviin, mitkä solut
olivat alkaneet ekspressoimaan transfektoitua plasmidia. Mikroskooppikuvia
analysoimalla saatiin varsinaisesti määritettyä ja laskettua kyseisten plasmidien
vaikutusta migraatioon ja solujen leviämiseen. Solujen kasvualustana käytettiin
peitinlaseja päällystettynä fibronektiinillä, jotta saatiin luotua tunnettu ja hallittu alusta.
Kokeiden avulla pystyttiin havaitsemaan kuinka FL Tal1:ä ekspressoivat solut
migroituvat ja leviävät hieman nopeammin kuin muita taliini1 –muotoja ekspressoivat
solut. Minitaliini4 -soluilla puolestaan havaittiin olevan hitain migroituminen ja heikoin
leviäminen muihin muokattuihin taliini1 –proteiineihin verrattuna. Kokeissa määritetyt
ja optimoidut laitteistot sekä olosuhteet antavat hyvän alustan jatkotutkimuksiin
fokaaliadheesioproteiinien parissa.
Avainsanat: fokaaliadheesio, taliini, solujen migraatio, solujen leviäminen
2
MASTER’S THESIS
Place:
University of Tampere
BioMediTech
Author:
Taina Viheriälä
Title:
Investigation of migration and spreading of cells expressing modified
talin1 forms
Pages:
61 pages
Supervisors: Docent Vesa Hytönen and MSc Rolle Rahikainen
Reviewers: Professor Markku Kulomaa and Docent Vesa Hytönen
Date:
August 2014
ABSTRACT
Focal adhesions (FA) are large complexes involving numerous proteins. Cells utilize
adhesions to interact with their environment and also with other cells. For example,
adhesions enable cells to move. One of the major proteins participating in the formation
and stabilization of adhesions is talin. It makes a link between a cell membrane protein
called integrin and the actin cytoskeleton which lies inside a cell. This enables a
response of a cell to the mechanical changes outside the cell. Many FA proteins have
been known for a long time but their features and functions are not yet fully known. In
this study we used truncated talin1 forms to see any differences in the migration and
spreading of the cells. At first, we had to optimize and set up the equipment for the
migration assay. Three cell lines were used in this part; wild type mouse embryonic
fibroblasts (WT MEF) and cells from which either talin or vinculin genes are removed.
In the actual migration assay talin1 knock-out cells were used but they were transfected
with the desired talin1 plasmid: full length talin1 (FL Tal1) or with one of the four
truncated talin1 forms. All the proteins are expressed with fusion to a red fluorescent
protein (mCherry) in order to see which cells have started to express the transfected
talin form. By analyzing the microscope images, the impact of the talin mutants on the
cells’ migration and spreading can be defined and measured. Fibronectin coated
coverslips were used in order to gain a known and controlled substrate for the cells to
grow. With the research we were able to observe how the FL Tal1 cells migrated and
spread faster than the other talin1 forms. Minitalin4 cells were assumed to be the one to
migrate and spread the slowest compared to the other talin1 forms used in this study.
This hypothesis turned out to be right. The equipment and conditions determined and
optimized give a good base for the future studies among focal adhesions proteins.
Keywords: focal adhesion, talin, cell migration, cell spreading
3
CONTENTS
ACKNOWLEDGEMENTS ............................................................................................. 1
TIIVISTELMÄ ................................................................................................................. 2
ABSTRACT ..................................................................................................................... 3
ABBREVIATIONS .......................................................................................................... 6
1.
REVIEW OF THE LITERATURE ........................................................................... 8
1.1
Introduction ........................................................................................................ 8
1.2
Extracellular matrix ........................................................................................... 9
1.3
Focal adhesions ................................................................................................ 10
1.3.1 Different types of focal adhesions .................................................................. 12
1.3.1.1 Non-integrin mediated focal adhesions ....................................................... 12
1.3.2 Formation of focal adhesions ......................................................................... 12
1.3.3 Maturation of FAs .......................................................................................... 14
1.3.4 Disassembly of focal adhesions ..................................................................... 14
1.3.5 Role of FAs in cell migration ......................................................................... 14
1.4
Proteins in FAs................................................................................................. 16
1.4.1 Talin1 ............................................................................................................. 16
1.4.2 Vinculin .......................................................................................................... 18
1.4.3 Integrins .......................................................................................................... 20
1.4.4 Talin interactions with other proteins............................................................. 22
1.5
Methods used in this project ............................................................................ 24
1.5.1 Scratch assay .................................................................................................. 24
2.
MAIN GOALS IN THE RESEARCH .................................................................... 26
3.
MATERIALS AND METHODS ............................................................................ 27
3.1 Mammalian cell culture ........................................................................................ 27
3.1.1 Cell strains ...................................................................................................... 27
3.1.2 Maintenance of the cells ................................................................................. 28
3.1.3 Fibronectin coated coverslips ......................................................................... 28
3.2 Migration assay ..................................................................................................... 29
3.2.1 Optimizing migration assay with scratch assay ............................................. 29
3.2.2 Optimizing migration assay with PDMS buttons ........................................... 31
3.3 Transfection of the cells ........................................................................................ 32
3.3.1 Plasmids used in the transfections .................................................................. 32
4
3.3.2 Immunostaining of the cells ........................................................................... 33
3.3.3 Antibodies used in this study ......................................................................... 34
3.3.4 Optimization of the electroporation ............................................................... 34
3.4 Microscopy ........................................................................................................... 35
3.5 Image analysis ....................................................................................................... 35
3.5.1 Image analysis in migration assays ................................................................ 35
3.5.2 Image analysis in migration assay using transfected Tal1-/- MEFs ................ 36
3.5.3 Image analysis in spreading assay .................................................................. 38
4.
RESULTS................................................................................................................ 40
4.1 Optimization of scratch assay ............................................................................... 40
4.1.1 Scratch assay used to study the migration of Tal1-/- MEF, Vin-/- MEF and WT
MEF cells ................................................................................................................ 42
4.1.2 PDMS buttons used to study the migration of Tal1-/- MEF, Vin-/- MEF and
WT MEF cells ......................................................................................................... 44
4.2 Migration study using cells expressing mCherry-tagged talin1 and truncated talin1
proteins........................................................................................................................ 45
4.3 Spreading assay using minitalin and full length talin plasmids ............................ 47
5.
DISCUSSION ......................................................................................................... 49
5.1 Optimization of migration assay ........................................................................... 49
5.2 Optimization of transfection ................................................................................. 51
5.3 Migration assay with transfected Tal1-/- MEF cells .............................................. 52
5.4 Spreading assay with transfected Tal1-/- MEF cells .............................................. 54
5.5 Cell culture substrates used in this study .............................................................. 56
6.
CONCLUSIONS ..................................................................................................... 58
REFERENCES ............................................................................................................... 59
5
ABBREVIATIONS
2D
3D
3T3
AB
ABS
BSA
DAPI
DDR1
DDR2
dH2O
DMEM
DS
ECM
EDTA
FA
FAK
FBS
FERM
FL
Fn
HUVEC
IBS
MD
MEF
MP
PBS
PDMS
PFA
PIP2
PTB
RHAMM
RT
Tal1
Tal2
TGF-β
TIRF
TLN1
TLN2
VBS
Vh
Vhd1
Two dimensional
Three dimensional
3-day transfer, inoculum 3 x 105 cells
Antibody
Actin binding site
Bovine serum albumin
4',6-diamidino-2-phenylindole
Discoidin domain receptor 1
Discoidin domain receptor 2
Distilled water
Dulbecco's Modified Eagle Medium
Dimerization domain
Extracellular matrix
Ethylenediaminetetraacetic acid
Focal adhesion
Focal adhesion kinase
Fetal bovine serum
Protein 4.1, ezrin, radixin, moesin
Full length
Fibronectin
Human umbilical vein endothelial cells
Integrin binding site
Membrane distal
Mouse embryonic fibroblast
Membrane proximal
Phosphate buffer saline
Polydimethylsiloxane
Paraformaldehyde
Phosphatidylinositol 4,5-bisphosphate
Phosphotyrosine binding
Hyaluronan-mediated motility receptor
Room temperature
Talin1
Talin2
Transforming growth factor beta
Total Internal Reflection Fluorescence Microscopy
Talin1 gene
Talin2 gene
Vinculin binding site
Vinculin head
Vinculin head domain 1
6
Vin
vWF
WT MEF
Vinculin
von Willebrand factor
Wild type mouse embryonic fibroblast
7
1. REVIEW OF THE LITERATURE
1.1 Introduction
Focal adhesions (FA) are involved in cell-extracellular matrix (ECM) interactions with
multiple different proteins with different functions. The most important proteins
involved in FAs are integrin, talin, vinculin and F-actin (Lee, Kamm & Mofrad 2007).
Integrin mediates the FAs by functioning as a transmembrane protein linking the ECM
and the actin cytoskeleton of the cell together (Anthis, Campbell 2011). Integrin has
binding sites for some proteins out of which talin works as a linker protein between
integrin and the cytoskeleton. In addition, talin has binding sites for vinculin which
again has binding sites for F-actin (Lee, Kamm & Mofrad 2007). Vinculin functions in
the adhesion site as a stabilizer, and it reinforces the binding of talin and F-actin (Golji,
Mofrad 2010). Among the linker feature talin can also activate integrins and vinculins
and make integrins to form clusters in the adhesion sites (Roberts, Critchley 2009).
Cells use FAs to aid in their migration, proliferation and to their motility. There can
exist different types of FAs. They can be small or large and also on different parts on
the cell depending on the cell and the ECM (Geiger, Yamada 2011). In order for a cell
to migrate it needs to form focal adhesions and then break them and repeat this cycle
over and over. The formation and the breakage of the adhesions must happen as a
continuum and in equilibrium in order for a cell to migrate. Integrin activation has a key
role in formation of the focal adhesions. It can occur in two ways: inside-out or outsidein. Talin is involved in the inside-out activation and ECM proteins are involved in the
outside-in activation (Anthis, Campbell 2011).
Talin is a 270 kDa cytoplasmic protein which has binding sites for integrin, vinculin and
F-actin. There are two talin genes in mammalian genome, TLN1 and TLN2, and they
encode for quite similar proteins talin1 and talin2, respectively. Talin1 is a more
commonly expressed and more thoroughly studied protein than talin2. This study also is
focused more on talin1. It consists of a globular head domain and a flexible rod domain
(Figure 3). Head domain again consists of four subdomains; F0-F3 from which the F3
subdomain is involved in the binding of integrin. Other binding sites for integrin and
also for vinculin are located in the rod domain and especially in its helical bundles
which require to be unfolded in order for the binding sites to be accessible for vinculin
(Roberts, Critchley 2009).
8
1.2 Extracellular matrix
Extracellular matrix (ECM) offers support and a signaling pathway for cells and it also
connects the adjacent cells together. There are numerous different types of proteins in
the ECM (Table 1), each with specific roles in the function of ECM, but some structural
proteins are perhaps more important than others. The pattern of ECM proteins can differ
between tissues: for example, collagens are the major structural proteins in connective
tissue. Fibronectin (Fn) is also a certain type of structural protein. It is a glycoprotein
which links collagen and integrin together so that cell adhesion and movement is
enabled (Frantz, Stewart & Weaver 2010).
Table 1. Components of the ECM. (Frantz, Stewart & Weaver 2010)
Component
Type
Location
White
connective
tissue
Yellow
connective
tissue
Collagen
Protein
Elastin
Protein
Keratin
Protein
Nails and hair
Laminin
Protein
Basal lamina
Function
Mechanical support, structural
framework
Impart elasticity to ECM
Mechanical support, structural
framework
Anchors cell surface to basal
lamina
Loose
connective
Cell adhesion, migration
tissue
Nidogen
Protein
Basal lamina Links laminin and type IV collagen
Fibroblasts and Angiogenesis, apoptosis, activation
Thrombospondin
Protein
endothelial cells of TGF-β and Immune regulation
Embryonic
Tenascin
Protein
tissues, wound
Role in development
and tumors
Circulation and Regulatory linker in cell adhesion
Vitronectin
Protein
different tissues
and cell invasion
von Willebrand
Platelet adhesion, blood
Protein
Circulation
factor (vWF)
coagulation
Cartilage,
Aggrecan
Proteoglycan
Mechanical support
Chondrocyte
Widespread in Binds to type I collagen fibrils and
Decorin
Proteoglycan
ECM
TGF-β
Structural and filtering function in
Perlecan
Proteoglycan Basal lamina
basal lamina
Fibronectin
Protein
9
1.3 Focal adhesions
Focal adhesions are cellular substructures making cells capable to interact and attach to
the extracellular matrix (ECM). In this way the ECM can also send information to the
cells so that focal adhesions are actually bidirectional communication channels. Cells
use focal adhesions to communicate with their environment and to make the right
responses like migration, spreading, proliferation and gene expression (Lo 2006). Focal
adhesions are thus very important players in many cellular events and they can also be
connected to some disease cases like cancer invasion and metastasis (Geiger, Yamada
2011). Previous studies have shown that cells make tight connections to the matrix at
focal adhesion sites leaving only a ∼10-15 nm gap between the cell membrane and the
substrate surface (Zamir, Geiger 2001).
There are numerous proteins, at least 180, involved in focal adhesions and they form
together a complex called adhesome (Geiger, Yamada 2011). From these proteins at
least 90 has been identified to locate directly at the adhesion site (Roberts, Critchley
2009). One of these proteins is the most important and it is called integrin (Lo 2006). It
is a membrane protein which can mediate focal adhesions together with other proteins
(Geiger, Yamada 2011). Integrin forms the direct link between cell and ECM
environment (Figure 1) (Anthis, Campbell 2011). Integrins are the only focal adhesion
proteins located in the ECM environment and there are several other proteins that are
constituting the focal adhesions inside the cell. It has been identified that there are at
least 42 proteins which bind directly to integrin (Rao, Winter 2009)(Roberts, Critchley
2009). Talin and vinculin are important proteins in the cytoplasmic side of the FA and
they are among the proteins forming the linking between integrin and F-actin (Lee,
Kamm & Mofrad 2007). They are also called integrin-binding proteins (Cohen et al.
2006). Cells lacking talin are shown to be unable to form focal adhesions as well as
show cell spreading (Lee, Kamm & Mofrad 2007). Altogether, these proteins form sort
of a linking cascade: ECM-integrin-talin-vinculin-F-actin (Figure 1) (Lee, Kamm &
Mofrad 2007). All the proteins that take part in focal adhesions can be divided into three
groups depending on their positions: extracellular, transmembrane or cytoplasmic and
into three subclasses depending on their function: integrin-binding proteins, adaptors
or/and scaffolding proteins and enzymes (Berrier, Yamada 2007). Integrins belong in
the transmembrane protein group and talin belongs to the cytoplasmic group together
with F-actin (Lo 2006). According to the protein function, talin belongs to integrin10
binding proteins, because it binds directly to integrin’s cytoplasmic domain and
therefore regulates the activation of integrin (Berrier, Yamada 2007). As the name
suggests, adaptor/scaffolding proteins like vinculin and paxillin in turn link the integrinbinding proteins to other proteins (Berrier, Yamada 2007). Focal adhesion also needs a
number of other enzymes that can for example phosphorylate or dephosphorylate
proteins to modulate their binding properties (Anthis, Campbell 2011). All proteins
involved in FAs either form a direct link to another FA protein (scaffolding interactions)
or modify another FA protein (regulatory interactions) so that it can take part into
adhesion (Geiger, Yamada 2011). Some proteins like talin can also interact with
phospholipids (PIP2) located in the cell membrane (Wehrle-Haller 2012).
Figure 1. Figure illustrating the major proteins in the focal adhesion site. Figure adopted from (Rao,
Winter 2009)
Scaffolding interactions eventually connect integrin tails to actin cytoskeleton. They
consist of actin-associated molecules and adaptor proteins. The first mentioned
scaffolding proteins, e.g. filamin, talin and α-actinin, interact directly with actin and
regulate its organization. The second protein subclass, adaptor proteins, can then interact
with the actin-associated proteins, with each other and with integrin (Geiger, Yamada
2011).
11
1.3.1 Different types of focal adhesions
Different types of focal adhesions can occur even in a same cell and they can be
morphologically, molecularly and dynamically diverse. There are a few types of
different cell adhesions that can occur in for example a migrating or spreading cell.
Classical focal adhesions (1.) are the most common types, they usually appear in a flat
surface and their size is about a few square micrometers (Geiger, Yamada 2011). The
second type (2.) of adhesions is dotlike nascent adhesions and focal complexes which
usually provide the first and early attachment sites (Berrier, Yamada 2007, Geiger,
Yamada 2011). They don’t survive as long as the classical form of adhesion but once
stabilized they can transform themselves into the classical focal adhesions. The third
type (3.) is elongated fibrillar adhesions, mainly located in the central parts of the cell.
The last type (4.) of the adhesions are podosomes or invadopodia which are ringlike
adhesions formed around actins (Geiger, Yamada 2011).
Usually cells utilize only one kind of adhesion type at a time. However when cells are,
for example, rapidly migrating they can sometimes form different types of adhesions at
the same time (Geiger, Yamada 2011).
1.3.1.1 Non-integrin mediated focal adhesions
It is good to note that not all interactions in adhesions are mediated by integrin. There
are few known receptors that can play an important role in cell-matrix interactions like
tyrosine kinase, which is one example of the alternative ECM receptors. It can make
interactions with collagen and then take part in signaling and adhesion etc. In addition,
DDR1 (discoidin domain receptor 1) and DDR2 receptors can function as integrin
independent receptors and CD44 and RHAMM are receptors that can function without
integrins in adhesion sites (Geiger, Yamada 2011).
1.3.2 Formation of focal adhesions
Cells can form adhesions differently depending on the type of the cell and the ECM.
Adhesions can be small or large and they can be unevenly distributed within the cell.
For example, there may be few small adhesions near the edge of the cell and large
adhesions located in the central region of the cell (Huttenlocher, Horwitz 2011).
12
Some proteins involved in adhesions exist in two conformational stages (e.g. talin and
vinculin); open and closed forms. Typically, open conformation represents the active
form and it enables ligand binding, and closed conformation is inactive. In order to
stabilize the adhesion the closed conformation needs to be switched to the open form.
Closed form is usually maintained with interactions within the protein itself. Therefore
the conformational change requires changes in the protein folding (Geiger, Yamada
2011).
The formation of focal adhesion sites starts from integrin activation. This can happen
through outside-in or inside-out activating processes. In the outside-in activation,
proteins in the ECM cause activation and in the inside-out activation, proteins inside the
cell cause the activation, respectively. After integrin activation, integrins start to cluster
which is triggered by the binding of cytosolic proteins to the integrin tail domains
(Humphries et al. 2007). The types of integrins involved in the FAs can change during
the adhesion. For example αvβ3 and α5β1 take part in formation of adhesions and
signaling, but their presence in matured adhesions is substrate-dependent (Geiger,
Yamada 2011). Activated integrins then allow talin to form a connection between talin
and F-actin. This connection is reinforced by vinculin. When these proteins are joined
together they form the initial complex of adhesion. The next step is to recruit other
proteins to gain stable and functional focal adhesion complexes connecting cell and
extracellular matrix (Golji, Mofrad 2010).
1.3.2.1 Chemical and physical sensing of ECM
Cells can sense ECM through chemical and also physical properties. A variety of
receptors, e.g. integrins, are included in chemical sensing. Receptors can have different
binding properties for specific ECM molecules such as collagen. These types of specific
connections are the key for adhesion formation. The requirements for the components
enabling formation of stable adhesions vary between different cell types. Integrin
receptors function together with growth factor receptors to sense the properties of the
ECM (Geiger, Yamada 2011).
Cells can also sense physical properties of the ECM like stiffness and ligand spacing.
Although physical sensing is not yet as well understood as chemical sensing, it is known
13
that the geometry and topography are important for the cell to sense the ECM (Geiger,
Yamada 2011).
1.3.3 Maturation of FAs
As previously described, in order to gain a stable and functional FA complex the initial
focal complex needs to be changed in terms of molecular composition and
rearrangement. This requires conformational changes in proteins and also changes in the
assembly of the proteins. An additional step in the maturation is the formation of
fibrillar adhesions. They are sometimes required, especially in fibroblasts, in the
maturation of integrin-mediated FAs. This process is force-dependent as is the
recognition of ECM in the assembly of focal complexes and focal adhesions (Geiger,
Yamada 2011).
1.3.4 Disassembly of focal adhesions
Disassembly of focal adhesions is an important step for cells for example during cell
migration. Cells need physical and biochemical interactions for releasing the connection
between ECM and integrins (Petit, Thiery 2000). Integrins can either be physically
broken off the membrane, they can form aggregates and move to the next adhesion site
or they can stay in the cell membrane and spread along it (Palecek et al. 1996).
Disassembly can occur also via a protein called calpain which can cleave the adhesion
components (Geiger, Yamada 2011).
Some ECM proteins have been shown to stimulate the disassembly of the focal
adhesion during the cell migration. A couple of these proteins are for example
thrombospondin and tenascin (Sage, Bornstein 1991).
1.3.5 Role of FAs in cell migration
Cell migration is a complicated process which requires co-operation between adhesion
proteins, signaling proteins and many other proteins. Cells need to form adhesions and
then break the adhesions and again form novel adhesions to enable cellular movement
(Figure 2). The best speed for migration is achieved when the formation and the release
14
of the adhesions are proficient. Therefore the rate for migration depends on how fast the
cell can form and break the adhesions. When a cell moves it pushes its fore, central and
back regions forward and this consumes lots of energy (Huttenlocher, Horwitz 2011).
The migration needs reinforcement of cell polarity to create leading edge and trailing
edge. The leading edge protrudes and new cell-matrix contacts are created while the
trailing edge retracts, causing the cell to move forward (Berrier, Yamada 2007).
Adhesion dynamics which constitute quite a wide range of mechanisms are an important
part in cell migration. Dynamics can be regulated by microtubules by destabilizing FAs
by releasing the tension. Signaling proteins in the initial complex can be bound very
transiently while the scaffolding proteins are making more stable interactions (Geiger,
Yamada 2011).
Intgerin-mediated migration has been shown to take part in a variety of human
development processes like wound healing, and it is also important for diseases such as
cancer (Huttenlocher, Horwitz 2011).
Figure 2. Picture of migrating cell shows formation and disassembly of FAs. Time elapse between the
two stages of FA is 5 minutes; green color indicates the first picture and red color the second picture.
Yellow color shows the FAs which were at the same spot in both pictures. Figure adopted from (Carisey,
Ballestrem 2011)
15
1.4 Proteins in FAs
1.4.1 Talin1
Talin is a 270 kDa cytoplasmic protein consisting of N-terminal head ( 50 kDa) and Cterminal rod ( 220 kDa) (Figure 3) (Roberts, Critchley 2009). Rod is composed of
62 helices that form 13 (R1-R13) amphipathic 4- or 5-helix bundles (Anthis et al. 2010).
Talin has 2541 amino acid residues with the head domain positioned in residues 1-400
and the rod domain in residues 482–2541 (Roberts, Critchley 2009). There are two talin
isoforms in the talin protein family; talin1 and talin2, but talin1 is expressed more
commonly than talin2 (Lo 2006).
Talin head contains FERM (residues 1-400) domain which in turn comprises of four
subdomains (F0, F1, F2 and F3) (Critchley 2005). The F0 subdomain is not always
counted as one of the FERM subdomains because it is not present in the typical FERM
protein (Anthis, Campbell 2011). FERM (protein 4.1, ezrin, radixin, moesin) domains
are found also in other proteins and they are often associated with biological processes
such as cell adhesions and proliferation (Elliott et al. 2010). The shape of FERM domain
(Figure 4) is usually cloverleaf, but in the case of talin it is found to be a linear
arrangement of subdomains (Anthis, Campbell 2011). Rod and head domains take part
in forming the globular and extended conformations of talin. When the rod domain is
isolated, it shows extended conformation, suggesting that forming globular
conformation requires interactions between the rod and the head domains (Goult et al.
2013). Rod domain is also responsible for the dimerization of the talin (Critchley 2005).
Figure 3. The schematic structure of talin with amino acid numbers. 62 α-helices are shown as ovals.
Orange ovals = vinculin binding sites (VBS). ABS = actin binding site, IBS = integrin binding site and
DS = dimerization domain. Figure adopted from (Wang et al. 2011)
The main function for talin is to participate to the integrin-mediated events and to act as
a linker between integrin and cytoskeleton. Talin also activates integrins and vinculins
whose activation is an essential process in the formation of FAs (Critchley 2005). In
addition talin promotes integrin clustering and also the switch of integrins from lowaffinity state to high-affinity state. But it is good to note that talin does not make all this
16
happen by itself, and it needs a co-operation with other proteins as well. Talin can also
be said to be a recruitment protein, meaning that it recruits other proteins to join the
focal adhesion (Roberts, Critchley 2009).
F0
F1
F2
F3
Figure 4. Crystal structures of one of the rod fragments (A) and FERM domains (B) of the talin.
FERM domain contains subdomains F0-F3 (F0 is not always counted as one of the subdomains). F1
subdomain contains five β-sheets with α-helix running across it. The F2 subdomain is completely αhelical with a short linker region. The F3 subdomain is composed of two β-sheets followed by an α-helix
(Calderwood et al. 2002). PDB ID 3DYJ from Gingras et al. (2009) and 3IVF from Elliott et al. (2010).
1.4.1.1 Talin2
Tln2 gene is another talin gene in the human genome (Praekelt et al. 2012). Talin1
functions in cell adhesion whereas the function of talin2 is still unclear (Debrand et al.
2009). Talin2 is expressed primarily in striated muscle and in the brain while talin1 is
expressed widely in different tissues (Anthis, Campbell 2011). Talin2 also has rod and
head domains as talin1 and they have sequence similarity of 74% so they encode very
similar proteins (Praekelt et al. 2012). Talin2 gene (190 kb) is an approximately six
times larger protein in mammals as talin1 (30 kb) and because of that talin2 also has
larger introns. Due to the larger size of talin2 gene there are also more splice variants in
talin2 gene and therefore it is possible that talin2 has some special functions in specific
tissues (Lo 2006).
Talin2 and talin1 have different affinities for integrins and for some reason talin2 has
higher affinity for β-integrin than talin1 (Wehrle-Haller 2012). They both have their
17
roles in FAs but their location is different; talin1 has shown to localize in peripheral
focal adhesions while talin2 participates in both focal and fibrillar adhesions (Praekelt et
al. 2012).
Lacking talin2 gene does not have as much of an impact for cells as lacking talin1 gene.
Although when considering single cells the lack of talin1 or talin2 does not show a huge
difference in their behavior but the differences become significant in the whole
organism (Praekelt et al. 2012).
1.4.2 Vinculin
Vinculin is a cytoplasmic protein (116 kDa) which is associated with focal adhesions in
the cytoplasmic site and contributes to the cell adhesion and motility (Carisey,
Ballestrem 2011). Vinculin and talin are both among the proteins essential in formation
of stable focal adhesions (Golji, Mofrad 2010). Vinculin mainly functions as stabilizing
cell-ECM junctions by participating in linking integrins to the actin cytoskeleton
(Debrand et al. 2009). Vinculin contains five different domains (D1, D2, D3, D4 and
tail). Domains 1-4 constitute a globular head (Vh) (residues 1-850) of vinculin. Vinculin
has also a third main domain besides the tail (residues 879-1066) and head and it is a
proline-rich neck domain (residues 851-878) which connects head and tail domains
together. Vinculin contains several binding sites for talin and F-actin. In domain D1 for
example is a Vh1 subdomain which can bind to talins vinculin-binding site (VBS)
(Golji, Mofrad 2010). Head domain can also bind F-actin. Tail domain in turn can bind a
protein called paxillin, F-actin and also some other proteins (Lee et al. 2008).
The presence of vinculin is not as important for the formation of focal adhesions as is
the presence of talin. Cells lacking vinculin can form focal adhesions, but cell spreading
and cell motility are not as efficient as in cells that express vinculin (Lee et al. 2008). In
vinculin deficient cells the integrin linkage to cytoskeleton is less stable and it is easily
broken by mechanical force (Cohen et al. 2006). However, Xu, Baribault & Adamson
(1998) published that lack of vinculin shows for some reason faster migration than the
wild type cells. Lack of vinculin is also shown to be associated with cardiac diseases
and with brain abnormalities (Cohen et al. 2005).
18
1.4.2.1 Vinculin exists in two conformations
Vinculin can exist in active and also in inactive forms (Figure 5) from which only the
active form is capable working in FAs. When vinculin is in inactive form, which is
actually an autoinhibitory conformation, it prevents vinculin interactions with proteins.
Head, especially D1 domain, and tail domains make intramolecular connections to each
other closing the binding sites for talin and F-actin so that vinculin can autoinhibit the
talin-vinculin interactions and also F-actin-vinculin interactions. Therefore, in order for
vinculin to be bound to talin, its conformation must be changed from autoinhitory to
active. Vinculin can be activated by binding to talin and especially three VBSs are
showed to have the capability to bind vinculin at this point. These VBSs are exposed to
vinculin by mechanical force during talin activation and therefore it can be said that
external force is needed in vinculin activation (Golji, Mofrad 2010). When vinculin is
activated by this so-called force-dependent pathway it may lead at the same time to
reinforcement of the actual focal adhesion (Lee et al. 2008). There is also another way to
gain activated vinculin and it needs both talin and F-actin interactions with vinculin
(Golji, Mofrad 2010).
Figure 5. Picture illustrating inactive and active forms of vinculin. Domains D1-D4 constitute the
head domain, between the D4 and D5 is the proline-rich neck domain and D5 is the tail domain. In the
active state are the binding sites for many proteins accessible (shown in the picture) and some of these
proteins (talin, α-actinin, F-actin and PIP2) can be involved in activating the vinculin. Figure modified
from (Carisey, Ballestrem 2011)
19
1.4.3 Integrins
Integrins are large heterodimeric adhesion receptors which connect the intracellular and
extracellular components. They are found in the cell membrane and they usually take
part in focal adhesion sites but are found to be associated also with other cell surface
receptors like growth factor receptors (Berrier, Yamada 2007). Integrins have
cytoplasmic and extracellular domains and also a transmembrane helix; intracellular
domains are shorter, approximately 20-50 residues, than extracellular domains (7001000 residues) (Wegener et al. 2007). Because of these domains integrins can mediate
cell-cell and also cell-ECM interactions (Roberts, Critchley 2009). Cytoplasmic domain
consists of tail-regions (one α and one β region) which again consist of membrane-distal
(MD) and membrane-proximal (MP) helices (Anthis et al. 2010). There are 18 different
α and eight different β subunits in the human integrin family and they can combine to at
least 24 different heterodimers differing from their ligand binding affinity (Anthis,
Campbell 2011). Integrins even in the same cell can function with different affinities
due to their numerous combinations (Geiger, Yamada 2011). Integrin α-subunit usually
contains around 1000 amino acids and β-subunit contains around 700 amino acids but
their size may vary (Anthis, Campbell 2011). Focal adhesions are not usually limited to
a certain type of integrin receptor although exceptions occur. Also recruitment of other
proteins is not usually limited in focal adhesions but in some cases only certain proteins
like a protein called tensin is recruited only on some focal adhesion sites leading to the
conclusion that there might be signaling specificity (Berrier, Yamada 2007).
Integrins, especially β cytoplasmic domains, form clusters in the membrane when
assembling focal adhesions (Petit, Thiery 2000). This clustering activates many nonreceptor proteins like focal adhesion kinase (FAK), serine/threonine kinases and lipid
kinases. Usually the activation of these and other kinases causes the recruitment of
proteins to adhesion sites (Berrier, Yamada 2007). Integrin uses its α-domain in focal
adhesions to interact with ECM and it uses β-domain to interact with talin (Petit, Thiery
2000). In focal adhesions and in migration, integrins play a major role by linking ECM
to actin cytoskeleton via talin (Anthis et al. 2010). Talin acts as a linker between
integrin and F-actin and it regulates the affinity of integrins for ligands (Praekelt et al.
2012, Anthis et al. 2010). All the proteins that interact with integrins β tail can be
classified into two groups: regulatory proteins and cytoskeletal proteins. Talin is an
20
example of a cytoskeletal protein. Numerous α and β subunits have been identified, both
individual and combined (Petit, Thiery 2000).
Each α and β chain of the integrin consists of extracellular domains, a single membranespanning helix, and a short cytoplasmic tail (Anthis et al. 2010). Thus integrins have
both extracellular and intracellular domains and they can respond to both extracellular
and intracellular signals, which is quite unusual for membrane proteins (Wehrle-Haller
2012). Integrins function as glycoprotein receptors in focal adhesions, they are actually
the major receptors in assembly and recognition of ECM and they facilitate cell
migration to the correct tissue location (Petit, Thiery 2000, Wehrle-Haller 2012). They
can also take part in some contacts between two cells and play a role in a variety of
other biological processes like wound healing and immune responses (Petit, Thiery
2000). On the other hand some integrin receptors have been found to link to certain
types of abnormal cell behavior like cancer. Expression of β6 integrin is normally
prevented in epithelial cells but in wound healing its expression is upregulated and
therefore, if by unfortunate incident the expression of β6 stays up, the cell can progress
into a cancer cell (Berrier, Yamada 2007).
1.4.3.1 Integrin activation
As transmembrane proteins, integrins work on both sides of the membrane, and their
activation can occur via inside-out or outside-in activation processes. When integrin
works in focal adhesions its conformation changes throughout the activation process. In
inside-out activation the affinity of ligands of ECM increases, finally leading to a step
where talin is bound to the cytoplasmic tail of the integrin. Outside-in activation starts
from binding to the ECM ligand and leads to intracellular changes leading to increased
cell viability, proliferation and growth (Anthis, Campbell 2011).
Considering integrin activation (Figure 6), the most important tail regions are the β1 and
β3 subunits. They also have a few splice variants with different activity affinity.
Different β-tails have also different MP regions which affects their activation (Anthis et
al. 2010). Cells that have mutations in their β1 and β3 chains have shown significant
reduction in forming focal adhesions (Lyman et al. 1997, Reszka, Hayashi & Horwitz
1992).
21
Figure 6. Picture illustrating the states of integrin. a) X-ray based extracellular structure with the
cytoplasmic tails. b) Intermediate conformation in the integrin activation process. c) Active state of the
integrin after ligand binding. d) Illustration of the integrin domains. E1-E4 are EGF-like domains and β-T
is the β–tail domain. Figure adopted from (Anthis, Campbell 2011).
1.4.4 Talin interactions with other proteins
Talin functions as a structural linker between integrins and actin cytoskeleton and is
therefore an important protein in cell adhesions (Lee, Kamm & Mofrad 2007). It can
also function as a regulator for other proteins to bind to integrins (Debrand et al. 2009).
Talin has binding sites for vinculin, actin and integrin (Anthis et al. 2010). Head domain
contains binding site for F-actin and β-integrin in its F3 domain. The second binding
site for β-integrin is located in the rod domain and there are also numerous binding sites
for vinculin and at least two binding sites for F-actin in the rod (Debrand et al. 2009).
R13 is the best characterized within the talin rod α-helix bundles and it interacts with Factin. It is located near the C-terminal end and talin can therefore offer a direct link
between the β-integrin tail and the actin cytoskeleton (Anthis, Campbell 2011, Gingras
et al. 2007). Rod domain is also responsible for maintaining talin in its inactive form
(Goult et al. 2009).
Vinculin binding sites (VBS) are located in the talin rod domain and they are consisting
of individual α-helices buried in α-helix bundles. In order for the VBSs to be accessible
for vinculin the bundles must unfold or at least structurally reorganize. The mechanism
for the unfolding is still poorly understood but one theory is that a mechanical force can
22
make the bundles unfold (Roberts, Critchley 2009). The opening of the bundles under
mechanical stress has been studied by using molecular dynamics simulations (Hytönen,
Vogel 2008).
1.4.4.1 Talin-integrin interaction
Talin can interact with a broad range of proteins. Talin-integrin interactions are
common in focal adhesions. Integrins can bind to rod and head domains of the talin
(Goult et al. 2010). As shown previously, integrins play a key role in cell adhesions and
talin takes part in integrin activation by interacting with the β-integrin tail (Anthis et al.
2010). In order for talin to bind integrin its integrin binding sites (IBS) must be
activated. Talin can form an autoinhibited form by making a linkage between the
F3 domain and a rod bundle (residues 1654-2344) so that it blocks the integrin binding
site (Roberts, Critchley 2009).
Talin activates integrins by interacting with β-integrin tails (Anthis et al. 2010). The
activation process requires talin F2-F3 domain pairs, but other talin domains are also
needed for full integrin activation (Critchley 2000). F0 and F1 domains play their own
role in this process (Elliott et al. 2010). F0 takes part in activating β3-integrin and also
enhances β1-integrin activation (Critchley 2000). Integrin activation is mediated via
interactions between β3-integrin NPxY motif and the talin F3 domain which has a
phosphotyrosine binding (PTB) fold ( nthis et al. 2010, arc a-Alvarez et al. 2003). F3
also binds to integrin membrane-proximal helix (Anthis, Campbell 2011). GarcíaAlvarez et al. (2003) showed in their research that β3-tail can be covalently attached to
talin F2-F3 domains and used that fusion protein in their research ( arc a-Alvarez et al.
2003). It has been revealed that amino acids with K324 residues in talin1 and D759 in
β-integrin also contribute to talin-integrin interaction (Anthis et al. 2009). Talin-integrin
interactions can be regulated by calpain proteolysis or through phosphorylation of either
talin or integrin (Calderwood et al. 2002).
When integrins are activated salt bridge between α- and β-integrin tails breaks at the
same time, which then affects the affinity state of the integrins for ECM ( arc aAlvarez et al. 2003, Tadokoro et al. 2003).
23
1.4.4.2 Talin-vinculin interaction
Vinculin (116 kDa) is an actin binding protein located at the adhesions sites (Roberts,
Critchley 2009). Talin has numerous binding sites for vinculin in its rod domain.
Vinculin can exist in two conformations; inactive and active. Inactive form is unable to
bind talin. In inactive form the head and tail domains of vinculin are in close interaction
with each other so that this blocks the binding site for talin. Talin itself and especially
the VBS is able to release the interaction between vinculin head and tail domains. The
VBS disrupts the inactive form by binding to vinculin head Vhd1 domain, and this
eventually leads to the point where the tail domain breaks apart from the head domain.
It is also suggested that the VBSs of talin are exposed only in the presence of Vhd1 of
vinculin so that VBS helices of talin would stay in closed conformation without the
stabilizing effect of vinculin itself (Cohen et al. 2006).
There are also other proteins and cell membrane lipids that can release the vinculin
autoinhibition, such as α-actinin, F-actin and PIP2. Talin is more likely to be the one
which breaks the head-tail interaction in vinculin because talin is usually co-localized
with vinculin rather than for example with α-actinin. Talin has also been found to be
one of the requirements for vinculin to localize in FA. Thus, talin-vinculin interactions
have a key role in regulating focal adhesion formation (Carisey, Ballestrem 2011).
1.5 Methods used in this project
1.5.1 Scratch assay
Scratch assay (or wound healing assay) is an in-expensive and widely used method to
measure cell migration in vitro. It is easy to use and doesn’t require lots of different
components. It basically works as follows: a scratch to a monolayer of cells is made and
it is then measured how fast the cells invade the scratch by migration. Measuring can be
made by looking at cells in different time points and taking pictures with a microscope.
This makes the assay very suitable for analyzing live cells instead of fixed ones. It is
important to take the first picture right after the scratch has been made in order to know
how wide the scratch was (Liang, Park & Guan 2007).
When cells are migrating on the surface they are thought to migrate towards a cell-less
space (Figure 7). In addition, when cells encounter another cell they can make
24
interactions with the cell and eventually change their migration direction. In order to
avoid cell contacts too fast, it is important to note that the monolayer should not be over
populated or in contrast there should not be too few cells (Liang, Park & Guan 2007).
1.5.1.1 In vitro scratch assay is a potential method to analyze cells in vivo
In vitro scratch assay is a capable tool for analyzing the cell behavior in vivo. By in vitro
assay one can also regulate the migration by affecting the cell interactions with ECM
and other cells. This method can be also used to detect fluorescent cells so it is capable
of measuring the migration of the transfected cells in comparison to non-transfected
ones (Liang, Park & Guan 2007).
There are also some disadvantages in the in vitro scratch assay method. It takes quite a
long time to fully get the results: first it takes a few days to grow the cell monolayer and
then it takes hours to days for the cells to close the scratch. It also consumes many cells
so this method is not optional for studies where the amount of the cells is limited
(Liang, Park & Guan 2007).
Figure 7. Figure illustrating the scratch assay method. The first picture indicates the time when the
scratch has been made (0 h). The second picture is after 4 hours and the third is after 8 hours from the
scratch making. The arrows are pointing to specific cells which are tracked to see how fast they migrate.
Figure adopted from (Liang, Park & Guan 2007).
25
2. MAIN GOALS IN THE RESEARCH
The main goal of this study was to set up methodology for studying how talin and its
different modified forms contribute to cell-ECM adhesions especially in terms of
migration and cell spreading. This can be modelled by using talin1 knockout (Tal1-/-)
mouse embryonic fibroblasts (MEF) and then transfecting the cells with different talin
constructs.
The project was basically divided into four steps: the first (1) step was to optimize the
migration assay; the second (2) step was to use the assay to study how talin affects the
migration. The third (3) step was to use the same cells to study cell spreading and the
fourth (4) step was to do the imaging analysis from the phase and fluorescence
microscope images. Transfection was done by electroporation and this step also needed
some optimization to know the best conditions to gain the highest possible transfection
rate. In the first step we didn’t use transfected cells but instead WT MEF, Tal1-/- MEF
and Vin-/- MEF cell lines.
26
3. MATERIALS AND METHODS
3.1 Mammalian cell culture
3.1.1 Cell strains
The cell strains were obtained as follows: mouse embryonic fibroblasts (MEF) from Dr.
Wolfgang Ziegler (Hannover, Germany), WT MEF and Vin-/- MEF cells were both
immortalized strains isolated from mouse embryos produced in heterozygous crosses of
Vin+/- mice (more details in Xu, Baribault & Adamson 1998). Tal1-/- MEFs were
isolated by gene disruption from embryonic stem cells (more details in Priddle et al.
1998) and they were kindly provided by Prof. Michael Sheetz (Columbia University).
WT MEF and Vin-/- cells were frozen in liquid nitrogen but Tal1-/- cells were previously
under maintenance of Rolle Rahikainen.
3.1.1.1 Minitalins
There are many different talin isoforms existing, from which, minitalins (Figure 8) are
shortened versions of the full length talin. Minitalins are designed so that certain
functional elements of the rod domain have been deleted without affecting its integrin
and actin binding.
Figure 8. Full length talin and minitalins 1-4. Blue circles = F0 + FERM domain, blue/green/orange
boxes = helix bundles of the rod domain, red bar = dimerization domain, red star = mCherry.
27
3.1.2 Maintenance of the cells
Cells were cultured in T75 cell culture flasks (Sarstedt). WT MEF and Vin-/- MEF cells
needed medium which contained 10% FBS (fetal bovine serum) (Lonza, BioWhittaker,
Cat. No.), 1% L-glutamate and 1% (v/v) β-mercaptoethanol. The rest of the medium
were DMEM with 4.5 g/l glucose (Dulbecco’s Modified Eagle Medium) (Lonza,
BioWhittaker® Cat. No. BE12-614F). β-mercaptoethanol was added immediately after
the cell culture. Tal1-/- MEFs required DMEM/F12 medium that contains L-glutamate.
The medium was prepared by adding 15% FBS. β-mercaptoethanol was not added to
Tal1-/- knockout cells. When the cell confluence reached approximately 70%, the cells
were divided, usually twice in a week. WT MEF and Vin-/- MEFs divided a little bit
faster than Tal1-/- MEFs. Subcultivation was done with preheated medium (37°C). The
cells were washed with PBS (phosphate buffer saline) and the cells were removed from
the bottom of the flask with trypsin-EDTA (Lonza, BioWhittaker, Cat. No. BE17-161E)
(200 mg/l). The subcultivation ratio was usually 1:9 or 1:12 but sometimes even as
small as 1:2, depending on when the cells were used for the experiments.
The cells which were stored in liquid nitrogen were thawed quickly to avoid cell death.
The cell vial was quickly moved to water heater (37°C), the suspension was transferred
to a centrifuge tube and was centrifuged (200 g, RT, 5 min) (Rotina 48R). The medium
was removed and discarded, and the cell pellet was resuspended to a new medium and
eventually transferred to a T75 flask.
Medium samples were taken for mycoplasma test to ensure that they were clean. This
was always done when a new vial was opened. An important thing also to note is that
between different cell types the laminar flow hood was cleaned with 70% ethanol so all
the cell lines were handled separately. This was done to avoid cross contamination
between the cells lines.
3.1.3 Fibronectin coated coverslips
Coverslips that were used in the migration and spreading tests were coated with
fibronectin isolated from human serum by gelatin affinity chromatography (Finnish Red
Cross, FFP8). The fibronectin concentration was 25 µg/ml when used. It was diluted in
PBS.
28
The coating occurred so that 40 µl of the fibronectin was added to a well on a 6-well
plate and a coverslip was placed on top of it. Then incubation was performed at 37ºC
for 1h followed by washing with PBS three times with fibronectin side upwards. At this
point the coverslips could be left on the plate immersed with PBS.
3.2 Migration assay
Migration assay was performed with two different methods: scratch assay (Figure 9)
and PDMS button assay (Figure 10). Scratch assay was made with previously published
protocol from Yanling Chen1 (2011). PDMS buttons were put into a 6-well plate before
adding the cells. Cells were cultured and the proper amount (2x106 or 4x106) was
transferred to a 6-well plate with 2-3 ml of medium. Cell counts were measured with
Cell Countess (Invitrogen). Cells were left to grow (18-24 h) to obtain proper
confluence (approximately 70%), after which the scratches were made or the PDMS
buttons were removed. In order to get rid of the dead cells and also cells that were
detached from the bottom, the wells were washed carefully with 2 ml of PBS and new
medium was added. Pictures were taken with a microscope (described later in more
detail) at three different time points: 0 h, 12 h and 24 h. The washing with PBS and the
medium change were also done before every picture taking.
3.2.1 Optimizing migration assay with scratch assay
There were a few variables to be optimized in the scratch assay: how many cells are
needed in each well, how the scratch is made, and what instrument can be or should be
used to make the scratch. We also wanted to find the optimal time points (0 h, 12 h and
24 h) to see any differences and also to make sure that the scratch is not completely
closed.
The first experiment was done to solve for the variables described above. It was done
with WT MEF cells by transferring 200 000 cells to each of six wells and 400 000 cells
to six different wells. The scratch was done with three different pipette tips to test which
is the most suitable size to make the scratch. The tip sizes were 10 µl, 200 µl and 1000
µl. Two scratches were made in each well. In half of these wells the normal medium for
1
http://www.bio-protocol.org/wenzhang.aspx?id=100; recessed 12.6.2013
29
the cells was used, and in the other half a starvation medium was used. The idea to use
starvation medium was to stop the cells from dividing and only allow them to migrate.
The scratches were imaged and saved as mosaic pictures of 1x3 images with an optical
microscope (Olympus IX71 with Surveyor software) (Table 2).
Figure 9. Scratch assay step by step. Step 1: Cells added to the plate and reached the proper confluence.
Step 2: Scratch is being made with pipette point. First picture (0 h) is taken. Step 3: Cells are starting to
migrate over the scratch area. The second picture (12 h) is taken. Step 4: Cells have almost closed the
scratch. The third picture (24 h) is taken.
Figure 10. PDMS assay step by step. Step 1: PDMS buttons are added to the plate and sterilized. Step 2:
Cells are added and they have reached the proper confluence. Step 3: PDMS buttons are removed. The
first picture (0 h) is taken. Step 4 and step 5: The cells are closing the button area. Second (12 h) and third
(24 h) pictures are taken.
Table 2. Settings of the microscope in the migration assay. Microscope which was in use was
Olympus IX71 with Surveyor software.
Objective
10x
Exposure
2.846 ms
Filter
1 (no filter)
Gain
1.80
Phase contrast
4
Offset
-1028
30
The second experiment was done with all three cell lines: WT MEF, Tal1-/- MEF and
Vin-/- MEF. This time the cell amount was only the 400 000 cells in each well, and the
use of 200 000 cells per well was abandoned.
3.2.2 Optimizing migration assay with PDMS buttons
Polydimethylsiloxane (PDMS) (Dow Crowning Sylgard 184) buttons were made by
mixing base (part A) and curing agent (part B) in ratio of 10:1 and pouring the mix in a
plate. Air bubbles arose to the surface and the biggest bubbles were removed manually
by using a pipette tip. The plates were incubated for 30-60 minutes in 57ºC and they
were left to cool for the next day. On the next day the actual buttons were made with
1.7 mm wide injection needle.
Five PDMS buttons were placed in each well from a 6-well plate and they were
sterilized with UV-light for 2 x 30 min. This step of the migration optimization was
performed with all the three cell lines: WT MEF, Tal1-/- MEF and Vin-/- MEF and the
cell amounts were the same as in the scratch assay. The cells were added after
sterilization and left to grow for about 24 hours. On the next day (or when the 70%
confluence was achieved) the buttons were removed with forceps. The wells were
washed with 2 ml of PBS and new medium was added. At this point pictures were taken
and also later at 12 h and 24 h after the removal of the buttons. The images were taken
with the same optical microscope (Olympus IX71 with Surveyor software) (Table 3) as
in the scratch assay but this time the images were mosaic images of 3x3 from the button
area.
Table 3. Microscope settings in the PDMS button assay. Microscope which was in use was Olympus
IX71 with Surveyor software.
Objective
10x
Exposure
7.161 ms
Filter
1 (no filter)
Gain
5.31
Phase constrast
4
Offset
-497
31
3.3 Transfection of the cells
In order to get the plasmids inside the cells and the cells to express the desired
construct, transfection was needed. It was done with electroporation which is based on
the voltage, time and pulses the machine gives. The reaction requires the cells, the DNA
(plasmid) and the Neon Transfection System (Invitrogen).
One transfection reaction was done with 1 million cells and 5 µg of DNA. The DNA
was in concentration of 500 ng/µl. The procedure was started with warming dishes
containing 10 ml of medium in an incubator. The cells and the plasmids were mixed in a
1.5 ml tube followed by transfer into an electroporation tube where the actual reaction
happened. Two types of buffers were also required for the reaction: buffer R and buffer
E2.
First the cells were pooled in a tube and the cell amount was measured with Cell
Countess. Then the proper amount of the cells (the amount depends on how many
reactions were done) were transferred to a 15 ml tube which was then centrifuged (RT,
200g, 5 min). Medium was removed very carefully. The pellet was washed lightly with
PBS and the centrifugation was repeated. The supernatant was removed properly. The
buffer R was added 100 µl/reaction and the cells were resuspended. Then the
resuspended cells were transferred to a transfection tube with the plasmid. The
transfection tube already contained the buffer E2. The electroporation was made with
the Neon Invitrogen with the optimized settings (see chapter 3.3.4). After that, the cells
were transferred to the pre-warmed plates and were left in the 37ºC incubator for 2448 hours.
3.3.1 Plasmids used in the transfections
The plasmids that were used in this project were full length (FL) Tal1-/- and minitalins
1-4. They all were tagged with mCherry which is a fluorescent protein. This was done
so that we were able to see which cells were expressing the plasmids (those which emit
the fluorescent dye) and which were not.
32
3.3.2 Immunostaining of the cells
This step was done after 24-48 hours from the transfection. It is important to note that
too long time will kill the cells because nutrients from the medium will eventually run
out. The procedure needed coverslips (24x24 mm, thickness: 1.5) to which the cells
were transferred. They were washed with 70% ethanol and dH2O before use to get rid of
any dust and grease. They were then left to dry in an air flow. The coverslips were
coated with fibronectin (see chapter 3.1.3); 40 µl (25 µg/ml) was added to each
coverslip.
The number of cells was calculated with Cell Countess and 100 000 cells were
transferred on top of coverslip which was immersed in 2 ml of medium. The cells were
then left to spread and migrate over certain amount of time. For most experiments, two
(0 h and 24 h) or three (0 h, 12 h and 24 h) time points were used.
After the time point was reached, the medium was taken away and the cells were
washed very carefully with PBS. The cells were now ready to be fixed. The fixing was
done with 4% paraformaldehyde (PFA) diluted in PBS which was added 2 ml/well and
it was left to influence for 10 min at 37ºC. The fixed cells were again washed with PBS
but twice this time. The cells could be left in PBS for a couple of days if necessary.
The cells were then permeabilized with 0.2% Triton x100 in PBS buffer for 5 minutes.
The buffer had to be heated to room temperature. After permeabilization the samples
were blocked for 20 minutes in RT in blocking buffer, which consists of 1% BSA,
5% FBS and 0.05% Triton in PBS. Blocking buffer was also used for dilution of
antibodies and for washing between treatments with primary and secondary antibodies.
Antibodies (AB) were used 25 µl/coverslip and incubated in RT for 1 h. After primary
antibody the coverslips were washed by dipping them into the blocking buffer,
incubating them in 40 µl of the same buffer and repeating that round twice. Secondary
antibody could then be added to each coverslip; 25 µl/coverslip and incubation at RT
for 1 h. The same wash round was done two times after incubation. The last step was to
dip the coverslips in water. The coverslips were then mounted in HardSet Mounting
Medium (Vectashield, Cat. No. H-1500) with 1/50 of the DNA stain 4',6-diamidino-2phenylindole (DAPI); 25 µl/coverslip. Samples were stored at +4˚C covered from light.
33
3.3.3 Antibodies used in this study
The antibodies used in the study are listed in Table 4. Primary antibody is supposed to
interact with the known target molecule while the secondary antibody attaches to the
primary antibody. Due to fluorescent labelling the secondary antibody gives the
fluorescent color to the fixed cells.
Table 4. Antibodies used in this study. *Praekelt et al. 2012
Antibody
mCherry
Vinculin
Talin2
Primary antibody
(dilution)
SICG (Sicgen) AB0040200
dilution (1:100)
Goat
Vin-AB (Sigma, Germany,
hVIN-1, #V9131)
dilution (1:1000)
Mouse
Tal2 68E7 (Cancer
Research Techonology)*
dilution (1:200)
Mouse
Secondary antibody
Color
(dilution)
Alexafluor-594
Red
dilution (1:500)
Chicken-anti-goat
Alexafluor-488
dilution (1:200)
Goat-anti-mouse
Green
Alexafluor-488
dilution (1:500)
Goat-anti-mouse
Green
3.3.4 Optimization of the electroporation
The first electroporation which was done in this project showed poor transfection
efficiency. Therefore, optimization of the electroporation parameters was performed.
The optimization happened in three settings which can be changed in the electroporation
machine: voltage, time and amount of the pulses. The first reaction was made with
settings of 1200 V, 30 ms and one pulse.
From these tests (Table 5), the second settings seemed to show a little better transfection
rate so those were decided to be the settings for the next trials.
3.3.4.1 Evaluating the efficiency of transfection
Transfection efficiency was evaluated by Surveyor microscope by comparing the
fluorescence channel and the phase channel. Another method which was used was to
34
use the fluorescence microscope on fixed and antibody stained cells. The microscope
images were taken and the evaluation was based on them.
Table 5. Electroporation settings tested.
Voltage (V)
Time (ms)
Pulses (amount)
1200 V
30 ms
1 pulse
1200 V
40 ms
1 pulse
1350 V
20 ms
2 pulses
3.4 Microscopy
In this study different kinds of microscopes were used. In the optimization of migration
assay the phase contrast implemented in the fluorescence microscope (Olympus IX71)
was used. It was also used in the trials not involving transfected cells. Fluorescence
microscope (AxioImager M2, Zeiss) with AxioCam using 40x oil immersion objective
(Zeiss Plan NeoFluar 40x / NA 1.3) (Table 6) or 20x oil immersion objective (Table 7)
was used with the transfected cells.
The former microscope contains Surveyor software which functions so that it is possible
to set certain points from the sample in advance to the microscope program. This allows
taking pictures from the same spots later on and measuring exactly how the cells have
migrated towards the scratch area.
3.5 Image analysis
All the images during the research were analyzed by using a program called ImageJ
(Rasband W., ImageJ, National Institute of Mental Health, USA). Different functions,
such as circularity and threshold, inside the program were used in order to get the
required information from the images.
3.5.1 Image analysis in migration assays
In the scratch assay the scratch areas were measured by calculating the widths of the
scratches by drawing a line between two cells on the opposite sides of the scratches.
35
The measurements were taken from six different points from one image (schematically
illustrated in Figure 11). The graphs were made by calculating the average widths and
standard deviations from every image and adding them to match the time points. This
was done to all the time points, and at the end there should be descending trend line
which tells that the cells have migrated towards the scratch area.
In the PDMS picture analysis the PDMS button areas were measured by calculating the
size of the whole cell-free area (Figure 12). There should also be a descending trend
which tells that the cells migrated towards the middle point of the button area.
Table 6. Fluorescence microscopy (AxioImager M2, Zeiss) with 40x oil immersion objective settings
in the spreading assay
Fluorescent dye
Channel
Exposure time
Exposure
Image
power
panel
mCherry
596
2s
100 %
eGFP
488
1s
100 %
DAPI
380
250 ms
50 %
7x7
Table 7. Fluorescence microscopy (AxioImager M2, Zeiss) with 20x oil immersion objective settings
in the migration assay.
Fluorescent dye
Channel
Exposure time
Exposure
Image
power
panel
mCherry
596
2s
100 %
eGFP
488
1,5 s
100 %
DAPI
380
250 ms
50 %
8x8
3.5.2 Image analysis in migration assay using transfected Tal1-/- MEFs
The pictures were taken from the fixed cells after 20 hours from the making of the
scratch so it was impossible to know anymore where the original scratch line was. The
average starting widths of the scratches were thus calculated from the optimization
experiments and that was thought to be a valid assumption for initial value.
36
The ImageJ software counted the number of the cells in the scratch area with a
“threshold” and an “analyze particles” command (Figure 13). The transfected cells were
counted by hand clicking every transfected cell using the “analyze particles” command.
Figure 11. Image demonstrating how the widths from the scratches were measured in all the time
points. The points where the widths of the scratches are measured are randomly selected. Scale bar is 200
μm.
Figure 12. Image demonstrating how the PDMS area was measured. Scale bar is 200 μm.
37
run("Crop");
run("8-bit");
run("Invert");
run("Brightness/Contrast...");
setAutoThreshold("Default");
//run("Threshold...");
run("Analyze Particles...", "size=100-Infinity pixel circularity=0.05-1.00 show=Nothing display
summarize");
Figure 13. ImageJ script used for calculating the total amount of the cells. The size was set to start
from 100 pixels to avoid cells in the edge of the image.
3.5.3 Image analysis in spreading assay
In spreading assay the cells were fixed at three different time points so that the
microscope pictures were also from three different points. The transfected cell outlines
were manually drawn and the circularity was measured [1] using ImageJ (Figure 14).
From these results the spreading of the cells could be evaluated. It is good to note that
only those cells which had one nucleus and were not surrounded by other cells were
analyzed so aggregates and cell clusters were left outside the analysis.
ImageJ gave the circularity in the form of 0-1: 1 is completely circular and 0 is highly
polarized. The cells were divided into three subclasses; round (circularity = 0.66-1.0),
half round (0.33-0.66) and spread (0-0.33).
[1]
38
run("Set Scale...", "distance=1 known=1 pixel=1 unit=px");
run("Line Width...", "line=10");
run("Colors...", "foreground=red background=white selection=yellow");
run("Set Measurements...", "area mean min centroid perimeter fit shape redirect=None
decimal=9");
//setTool("freehand");
run("Add to Manager");
roiManager("Add");
roiManager("Select", 0);
roiManager("Split");
roiManager("Delete");
roiManager("Select", newArray(0,1));
roiManager("Draw");run("Input/Output...", "jpeg=85 gif=-1 file=.csv use_file copy_column
copy_row save_column save_row");
String.copyResults();
String.copyResults();
run("Select None");
run("Close");
run("Flatten");
Figure 14. ImageJ script used for calculating the circularity of the cells.
39
4. RESULTS
4.1 Optimization of scratch assay
The first part of this project was to optimize the migration assay. It was done with two
methods: scratch assay and PDMS buttons. Three cell lines were used in both methods;
Tal1-/- MEF, Vin-/- MEF and WT MEF cells.
The first experiment was done only with the WT MEF cells. The parameters to optimize
were the amount of cells and the pipette tip size to be used. We also wanted to know if
the pre-selected time points (0 h, 12 h and 24 h) were chosen adequately. The first
experiment was done with 400 000 and 200 000 cells in one well (Figure 15) and pipette
tip sizes were 10 µl, 200 µl and 1000 µl (Figure 16). The cells were left to grow for 2224 hours in the medium. The optimal cell amount is important because the nutrients in
the medium will eventually run out, but also it is important so that there are enough
cells to be able to see a cell line without overcrowding the well. 200 000 cells showed to
be too few because when the scratches were made there were no clear cell lines so it
would be quite hard to measure how fast the cells invade the scratch. 400 000 cells on
the other hand showed to be a quite good amount; the boundary between populated and
empty areas was clearly visible, but the well was not too crowded.
A
B
Figure 15. Scratches in plates of 200 000 cells (A) and 400 000 cells (B). The edge of the area
containing attached cells is better visible in a well where 4x105 cells were plated as compared to a well
with 2x105 cells. Scale bars are 200 μm.
40
The second variable was the pipette tip size. In order to get rational time points the
10 µl
1000 µl
width of the scratch should not be too small because then the cells would invade the
scratch area too fast.
10 μl
200 μl
1000 μl
Figure 16. Scratches made with different pipette tip sizes. Left scratch is made with 10 µl, the middle
scratch is made with 200 µl and the right scratch is made with 1000 µl micropipette tips. Scale bar is 200
μm.
The scratch width made with 10 µl tip was found to be too small so that after 24 hours
the scratch had been completely vanished. The scratch width made with 1000 µl tip on
the other hand was good in size but when comparing the width with the scratch made
with 200 µl tip the widths were almost the same. That might be because it was harder to
keep the right angle between the surface and tip with the 1000 µl because it was quite
large, so it was easier to make the scratch with the tip size of 200 µl. Due to these facts
we ended up using the 200 µl pipette tips.
Starvation media was also used along with normal media to test if it is possible to get a
situation where the cells would migrate but not divide. An issue occurred during the
tests; the cell amounts should be larger in order to get good confluence in the proper
amount of time, so we decided not to use the starvation media.
41
4.1.1 Scratch assay used to study the migration of Tal1-/- MEF, Vin-/- MEF and WT
MEF cells
After the optimization part, it was concluded that the cell amount to be plated is 400 000
cells/well and the pipette tip size should be 200 µl. The time points were also set up to
be 0 h, 12 h and 24 h.
The next step was to study if there are any differences in the migration between all three
cell lines (Figure 18). They were all plated, and after making the scratch, the plates were
imaged and the images were analyzed.
WT MEF cells showed a clearly faster migration rate than Vin-/- and Tal1-/- cells
(Figure 17) although Vin-/- cells and Tal1-/- showed also a clear migration. P-values
comparing WT MEF cells between Tal1-/- or Vin-/- revealed a statistical significance in
the difference in their migration rates.
100%
90%
80%
70%
60%
TAL1 -/Scratch area 50%
VIN -/-
40%
WT MEF
30%
20%
10%
0%
0h
12h
24h
Figure 17. State of migration in time points 0 h, 12 h and 24 h. The widths of the scratches are
normalized in between the experiments. 100% indicates the width of the scratch in the beginning of the
experiment. The 0 h, 12 h and 24 h are the time points when the images were taken. Analysis with t-test
showed the WT MEFs migrating faster than Tal1-/- cells (12 h, p < 0.001; 24 h, p < 0.05). Analogously,
Vin-/- cells and WT MEF cells were found to have p-values (12 h and 24 h) < 0.001.
42
0h
12 h
24 h
Tal1-/- MEF
Vin-/- MEF
WT MEF
Figure 18. Migration assay for cell spreading kinetics analysis. All three cell lines and microscope
images from time points 0 h, 12 h and 24 h are shown. The widths of the scratches are getting smaller
during the time. Phase contrast microscopy (Olympus IX71) was used to take the images. Scale bar is 200
μm.
43
4.1.2 PDMS buttons used to study the migration of Tal1-/- MEF, Vin-/- MEF and
WT MEF cells
PDMS buttons were the other method used to study the migration of the cells. It was
used to see if it is a more compatible and easily repeatable method compared to the
scratch assay. PDMS tests were made with all the three cell lines. Time points in this
method were set up similarly to those in the scratch assay; 0 h, 12 h and 24 h.
WT MEF showed to migrate faster than Tal1-/- and Vin-/- cells (p < 0.001) (Figure 19) in
this method also. The differences between Vin-/- and Tal1-/- were quite big even when
the error bars are taken in and statistical analysis indicated p < 0.001. According to the
graph below, the differences seem to be quite good between the cell lines and might
include a statistical significance, but the problem was that the results were different in
all the three times this assay was performed (data not shown in this thesis) compared to
the scratch assay where the results were the same.
100%
90%
80%
70%
60%
TAL -/-
PDMS button area 50%
VIN -/-
40%
WT MEF
30%
20%
10%
0%
0h
12h
24h
Figure 19. State of migration speed in time points 0 h, 12 h and 24 h. 100% indicates the area of the
PDMS button and the areas measured are normalized according to the starting area. The graph tells how
much of the area the cells have occupied as compared to the starting area of the PDMS button. Analysis
with t-test showed the p-values for all the cell lines compared to each other to be < 0.001 so there is a
statistical significance between the cell lines.
The method using PDMS buttons was not chosen for the future studies because it
appeared not to be as repeatable as the scratch assay and also because the scratch assay
is a very commonly used method in migration studies.
44
4.2 Migration study using cells expressing mCherry-tagged talin1 and truncated
talin1 proteins
The actual migration study was done with Tal1-/- MEF cells by transfecting them with
plasmids encoding minitalin and FL (full length) talin. 3x106 cells were transferred to a
coverslip after about 24-48 hours from the transfection and they were left to grow on a
coverslip for about 24 hours. Then the scratches were made to each coverslip. The cells
were fixed after 20 hours of spreading and they were antibody stained with mCherry
(produced in goat) (color: red) and talin2 (produced in mouse) (color: green) primary
antibodies and Alexafluor-594 (anti-goat produced in chicken) and Alexafluor-488
(anti-mouse produced in goat) as secondary antibodies. The VectaShield included DAPI
dye so the third color (blue) indicates all the nuclei.
Two scratches were made on each coverslip and 2-4 microscope images were taken
from each of the scratches after 24 hours of incubation (Figure 21). One image was also
taken from an area outside the scratch. The number of the cells in the scratch area was
calculated and the results were compared with the value obtained from the outside of the
scratch area.
Figure 20. Transfected cells in the scratch area (left graph) and transfected cells outside of the
scratch area (right graph) after 24 hours of migration. The bars indicate the number of the transfected
cells which can also be considered as a measure of the transfection efficiency. n = the total number of the
cells in all microscope images analyzed. Standard deviations were not used in the right graph because
only one image was taken from the area outside of the scratch.
The transfection efficiency was about 10% or less (Figure 20) in all the cases.
Minitalin2 sample showed to have fewer transfected cells than the other samples so the
45
reason for small number of cells in the scratch area might be due to a poor transfection
rather than due to a slow migration.
Figure 22 shows a comparison of the number of transfected cells in the scratch versus
outside of the scratch and indicates that all the constructs except minitalin4 appeared to
have transfected cells concentrated in the scratch area. In the case of minitalin3 the
preference was minimal.
Figure 21. Immunofluorescence staining of the Tal1 -/- cells transfected with different plasmids.
Fluorescence microscopy (AxioImager M2, Zeiss) with AxioCam using 40x oil immersion objective
(Zeiss Plan NeoFluar 40x / NA 1.3) was used to image these cells. Three different channels were also
used: mCherry (red), vinculin (green) and DAPI (blue). The line in the middle of the images is the scratch
with is closing due to the migration of the cells. Time point for the fixing was 20 hours. Scale bars is 100
μm.
46
2,5
2
1,5
1
0,5
0
FL
Mini1
Mini2
Mini3
Mini4
Figure 22. Fold change comparing the number of transfected cells in the scratch area to the number
of transfected cells outside the scratch after 24 hours of migration.
4.3 Spreading assay using minitalin and full length talin plasmids
The spreading assay was done also with Tal1-/- MEF cells by transfecting them with
plasmids expressing for minitalin and FL talin. In this case 3x106 (20 min and 2 h time
points) or 2x106 (20 h time point) cells were transferred to a coverslip 24-48 hours after
transfection. The cells were fixed at time points of 20 minutes, 2 hours and 20 hours.
The cells were then antibody stained with mCherry and Talin2 primary antibodies and
Alexafluor-594 and Alexafluor-488 as secondary antibodies. The blue color (nuclei)
from DAPI in the VectaShield mounting medium also appears at this time.
Cell spreading was measured by drawing the cell outlines by hand and by using ImageJ
to calculate the circularity (values from 0 to 1). The results were divided into three
subclasses; spread, half round and round.
At the 20 minute time point most of the cells seemed to still be round (Figure 23 and
24). FL, minitalin1 and minitalin2 showed a little spreading. At 2 hours timepoint all the
transfected cells showed signs of spreading and cells transfected with minitalin2 and
minitalin3 plasmids showed more spreading than the others (Figure 24). Samples
transfected with minitalin3 and minitalin4 showed approximately the same amount of
the spread cells (40%) at 2 h but in the time point at 20 hours a difference was observed;
minitalin3-transfected cells had spread faster than those cells expressing minitalin4
(30% compared to 55%). 25% to 55% of the cells were spread after 20 hours and only
47
2% to 15% of the cells were round after 20 hours. Therefore, the number of spread cells
between the samples expressing different talin forms was more divergent than the
number of the round cells.
Figure 23. Cell spreading images in time points of 20 min, 2 h and 20 h. Transfected cell lines (those
which are counted in) are marked with red color. Scale bars is 50μm.
Figure 24. Graphs illustrating cell spreading in different time points. The spreading area of each cell
was defined by drawing it manually and the circularity was measured by using ImageJ. The cells were
divided into three subclasses (spread, half spread and round). In the right lower corner there are examples
of the cells from those three categories shown.
48
5. DISCUSSION
Comparative analysis between two migration assays was performed in this study:
scratch assay and PDMS button assay were evaluated for the performance and
repeatability. Spreading assay was also performed in this study. The ultimate goal of the
study was to evaluate the migration and spreading rates between cells expressing
different talin1 constructs (FL talin1 and four minitalins).
5.1 Optimization of migration assay
Our results from the optimization of the scratch assay showed that the migration assay
done this way is easily repeatable and image analysis can be done easily. The protocol
for this method was taken from Yanling Chen (2011) though only steps 1-8 were
performed. In the first part of the study we tested different methods for the preparation
of the cell-free area and tried also different media to test whether it is possible to block
the cell proliferation during the migration analysis. The starvation medium, containing
0.5% BSA, appeared to be not as good as we thought because it made the cells migrate
slower and it did not actually completely stop the cell division. We decided to reject the
starvation medium and continue only with the normal medium which included all of the
necessary nutrients.
The migration assay continued with the selected equipment for testing if there is any
difference in the migration between WT MEF, Tal1-/- MEF and Vin-/- MEF cells. Xu et
al. (1998) published that vinculin knockout cells show faster migration than their wild
type counterparts. In addition, vinculin is not necessary for the formation of FAs but it
will stabilize them, and Goult et al. (2010) published that overexpression of vinculin
might lead to a suppression of the migration. These two studies support each other,
suggesting inhibitory role for vinculin in cell migration. It appeared that we, however,
got results where the WT MEF cells migrated fastest out of the three cell lines used. WT
MEF cells express all the necessary proteins in order to form stable focal adhesions with
the ECM unlike the Tal1-/- MEF and Vin-/- MEF cells, which are lacking talin1 or
vinculin, respectively. It is known from previous studies (Albiges-Rizo et al. 1995, Lee
et al. 2007) that the lack of talin has an impact on formation of focal adhesions and that
the cells will show much lower degrees of migration and cell spreading. Vinculin in turn
is a protein which stabilizes adhesions and contributes to the linkage between talin and
49
F-actin. Therefore, the lack of vinculin might influence the adhesion turnover leading to
faster adhesion formation and disassembly which again might lead to a faster migration.
Therefore we were surprised to see that Vin-/- and Tal1-/- MEF cells migrated slower
than the WT MEF cells. As mentioned above, cells lacking talin1 are unable to form
normal talin1 consisting focal adhesions but the question is, why did the Tal1-/- cells
show migration at all? One possible answer is talin2 protein which compensates for the
lack of talin1. Talins have quite similar properties although it has been shown that the
lack of talin2 does not have as much of an impact on the focal adhesions as talin1
(Praekelt et al. 2012). This supports our results that migration could still occur although
talin1 was not present. Praekelt et al. (2012) also showed that talin1 was expressed in
the edge of the cells whereas talin2 was more commonly expressed in the middle region
of the cells.
P-values for the differences between averaged WT MEF and Tal1-/- cell migration states
were < 0.001 (12 h) and < 0.05 (24 h) so there is significant difference in the migration
between those two cell lines. P-values for the WT MEF and Vin-/- were both < 0.001 so
there is also a significant migration difference between those cell lines. Another
migration assay which was also performed (graph is not illustrated in this thesis) gave
the p-value of < 0.001 for all the cell lines.
PDMS buttons were also tested in migration assay to see if they fit better for this assay.
The buttons were made just for this trial in our own laboratory and they were easy to
prepare, did not consume too much time and were inexpensive. Establishment of buttonbased assay required only a little more work than that of the scratch assay but was still
reasonable. It was also thought that this method would be more repeatable due to size of
the button which would always be the same on the contrary to the width of the scratch
which can change due to so many variables. The assay was repeated a few times but the
results were inconsistent (data not shown in this thesis) every time. On that account we
concluded that PDMS assay is not repeatable with these arrangements. It should be also
noted that the two figures (Figure 11 and Figure 12) should not be compared directly to
each other. In case of the scratch assay (Figure 11), the width of the scratch is reported.
In contrast, the PDMS button analysis (Figure 12) reports how much of the button area
has been occupied by the migrating cells. Therefore, the results should be evaluated
only within one assay type.
50
The t-test analysis showed p-values for the differences between averaged WT MEFs and
Tal1-/- in the PDMS button analysis to be <0.001 in both time points. The p-values for
the differences between WT MEF and Vin-/- were also < 0.001 in both time points. The
assay was repeated (graph is not illustrated in this thesis) and the p-values were also
< 0.001 at this time.
PDMS buttons were easy to use and showed promising results at first. However, this
wasn’t enough for us to choose this method for the future experiments. A slight problem
occurred during this method which could explain why the results were different every
time. When the cells were plated and they started to reach the confluence, they migrated
and randomly collided with the PDMS button. This changed their direction of
migration. Then, when the PDMS buttons were removed and the cells should have
started to migrate towards the button area, we observed that they had formed kind of a
wall which slowed down the migration. It thus took certain time for the cells to change
their direction again. It is also possible that the cells coming from behind started to
migrate across the other cells. Either way, this was a problem when studying the
migration by this method. Another reason why we rejected this method was the situation
when cells had migrated towards the center and they were starting to lack space. The
cells start to run into each other which could eventually disrupt the free migration and
results may be distorted. For these reasons and also because the scratch assay is very
commonly used method in migration research this method was not selected for further
studies.
5.2 Optimization of transfection
There were adversities during the research in the transfection and more precisely in its
efficiency. This was however a crucial step in order to gain any information about
differences between different talin1 forms. The first experiment that was done to test the
transfection gave almost no transfected cells. With optimization, we managed to find
settings enabling transfection efficiency in range of 10% (Figure 20). Earlier
experiments (results not shown) had indicated that Tal1-/- MEF cells would not enable
excellent transfection efficiency, so we agreed that 10% would adequate in order to
continue for further studies (Table 5).
51
The transfected plasmids differed in size, FL talin1 was the longest (12336 bp) and the
minitalin4 was the shortest (6143 bp). It is obvious that it could influence the efficiency
of the transfection because the FL plasmid could have been too large for the cells to
effectively take it in. In addition to this there are other possible factors influencing the
transfection efficacy such as the plasmid purification method used.
There are also other ways to work with cells which are expressing the desired proteins
such as stable cell lines. This means of creating cell lines which are stable and which
continuously express the protein construct of interest. Although this could help to
overcome the challenges in the transfection, it would have been too time consuming to
produce them first.
5.3 Migration assay with transfected Tal1-/- MEF cells
Albegies-Rizo et al. (1995) published that cells where talin was down-regulated formed
adhesions slower than their wild type counterparts. Instead of measuring the adhesion
speed we measured if there were more transfected cells in the migration area (the
scratch area) or in the so-called static area (area outside of the scratch). This indeed tells
if the transfected cells are able to migrate more or less as compared to their
nontransfected counterparts.
In the actual research we tested if there was any difference in the migration velocity in
between cells transfected with different talin1 constructs. In addition, the speading of
the transfected cells were evaluated. All the talin forms studied here were tagged with
fluorescence marker mCherry which would help us to see which cells have started to
express the proteins encoded by the plasmid. Our hypothesis was that the FL talin1
would show perhaps faster migration than the other constructs due to the fact it still
contains all the necessary domains and the proteins were also intact. Minitalins on the
other hand are lacking some ligand-binding sites so that the fourth minitalin has only its
head domain and therefore is missing all the vinculin binding sites and also the integrin
and F-actin binding sites located in the rod domain. It was estimated that the cells
expressing minitalin4 would probably form fewer adhesions than the other constructs.
However, it was hard to predict how adhesion formation would correlate with migration
velocity.
52
Vinculin is not necessary for the formation of FAs but it stabilizes them. In addition, the
talin-vinculin ratio in an ideal focal adhesion complex would be 1:5 or even 1:10
(Critchley 2009). Therefore, when vinculin is present it might lead to a suppression of
the migration like Goult et al. published in 2010. This on the other hand suggests that
the minitalin4-expressing cells would not necessarily be the slowest one to migrate
although vinculin is present, because the cell is incapable of using it. The fact that
minitalin4 cells are unable to use vinculin by the talin1 could eventually disturb the cell
migration less than in cases of minitalin2 and minitalin3 where the vinculin is in use but
not normally.
The results obtained with transfected Tal1-/- cells from the migration assay were that the
samples treated with FL talin1 and minitalin1 and 2 appeared to have more transfected
cells in the scratch area than outside of the scratch, although the large error bars should
be taken into good consideration. Minitalin4-treated sample in turn showed clearly
fewer transfected cells in the scratch area. This supports our hypothesis that the
minitalin4-expressing cells would migrate less than the FL Tal1 –expressing cells.
It would have been advantageous if the talin1 plasmid transfected in the cells would be
the only talin gene inside the cells. As previously mentioned the TLN2 gene was still
present in the genome so we should not think about the results as being obtained only
with the transfected plasmid. Talin1 and talin2 have a lot of sequence similarity so it is
obvious that they participate in the same events. Praekelt et al. (2012) published that
they both take part in the migration but occur in different parts of the cells. Talin1 is
more commonly in the edge region whereas talin2 exists in the middle regions of the
cells. Therefore, the significance of talin1 in the cell migration is greater than talin2.
Although, in case of Tal1-/- cells there might occur significance chances in the role of
talin2.
We measured the migration in a whole cell population. Another way would have been
to study individual cells. This would have required a microscopy technique to track
desired cells in real time. This way the measurement would have been different; we
could have measured the speed of migration for individual cells but that would have
required a lot of data in order to generalize the results for the whole population. The
advantage for this is that the poor transfection efficiency would not have been such an
issue as it was now when we evaluated the whole cell population.
53
5.4 Spreading assay with transfected Tal1-/- MEF cells
The role of talin in the cells has been studied by down-regulating it. Albiges-Rizo et al.
(1995) published that spreading was reduced and the focal contacts were smaller when
the talin was down-regulated. In the spreading assay the hypothesis was the same as in
the migration assay; FL talin1 –expressing cells would spread more and perhaps faster
than the cells expressing minitalins, especially the minitalin4. Here we did measure how
fast the cells were spreading by measuring the size and shape of the cells in three time
points. In the time point of 20 minutes the FL talin1 and also somehow surprisingly the
minitalin2 construct had started to show some spreading already. Of course it is notable
that 20 minutes after the cells were transferred onto a coverslip they were still quite
fragile because the attachment to the ECM may have not completely occurred yet.
When the medium was taken away and the fixing solution PFA was added, there was a
risk that the cells might detach from the bottom, so this step required extra caution and
had to be done very slowly. Even with these precautions, the risk of the detachment was
not eliminated.
Talin can form an autoinhibited form by interaction between its head and tail domains.
Because minitalins 1-3 contain fewer tail bundles and mintalin4 does not contain any
bundles, their ability to form the autoinhibtion conformation may be disturbed. This
might lead to an early integrin activation which can explain why the minitalin2 started
to show some spreading already after 20 minutes. On the other hand, the lack of tail
bundles might complicate the formation of stable adhesions. In the second time point,
2 hours, the cells should have formed more stable adhesions and, due to that, the risk for
the cells detaching from the plate was not high anymore during PFA-medium exchange.
In 2 hours all the constructs had started to spread and in the classification we used, all
kinds of spread cells were found. Minitalin2 appeared to still be in the lead although it
was a quite minor lead. FL talin1 was actually showing fewer cells in the spread
subclass than the others. In 20 hours the cells transfected with FL talin1 construct has
caught up and bypassed the others and had almost no round cells. The minitalin4transfected cells appeared to have fewest spread cells and also the most round cells. Still
the major components in every construct were the spread cells. This could be due the
fact that the FL construct is able to form stable adhesions but the minitalins are possibly
active only at the beginning of the cellular spreading process, leading to a good start for
the spreading but not ending up as spread as in the case of the full length talin. If
54
considering only the last time point our hypothesis was right: FL talin1 –expressing
cells had spread more than the others and the minitalin4-expressing cells less than the
others. Minitalin4-transfected cells also appeared to show no change in the spread form
between the last two time points. The reason might be that there is nothing but the talin2
to stabilize the adhesions so the cells just randomly form adhesions and then
disassemble them. The cells still express the talin2 and it takes part in the formation of
adhesions. In the case of minitalin4, it might be that the minitalin is competing with
talin2 for the binding to integrin. This is supported by the publication from Lee et al.
(2008) who stated that cells lacking vinculin showed to have unstable and easily
breakable adhesions.
Zhang et al. (2008) reported that talin2 protein did not affect the initiation of the
spreading in the talin1-null cells but instead the initiation of the spreading was driven by
the integrin and F-actin. Therefore they stated that the talin is not necessary for the
initiation of the cell spreading. In addition to the initiation, the spreading occurred
normally when the talin2 was present and this could be due to the possibility that talin2
protein expression might have increased. They also reported that during Tal2-shRNA
treatment, the cells began to become round after 48 hours. Since our research ended
after only 20 hours, a longer experimental time might be a good step for the future
plans. It is good to note that they used Tal2 shRNA in their experiment while we did
not, so our results and their results should not be directly compared to each other. They
also published that cells lacking both talin1 and talin2 became round after 28 minutes,
so it is clear that our cells contained talin2 because there were no sign of the cells
becoming round again, at least not in 20 hours.
Although the minitalin4 did not show any significant increase in the number of spread
cells between 2 hours and 20 hours, at least the number of round-shaped cells decreased
(Figure 24). Zhang et al. has described the role of vinculin in the adhesions when the
cell was not expressing the talin. They did not observe the vinculin in the FA sites but
they saw it localizing in the cytoplasm. This is supported by the role of vinculin which
functions to link talin to F-actin. When there is no talin there is no need for vinculin. All
combined, they discovered that talin depletion caused loss of FA and also cell rounding
at later times, but the lack of talin did not however prevent or delay the initiation of the
spreading. This is supported by the publication of Elliott et al. (2010) who used human
umbilical vein endothelial cells (HUVECs) which are normally already lacking talin2.
55
They published that lack of both talins showed reduced spreading and FA assembly and
only fewer than 50 % of those cells were spread after 24 hours. They used cells
expressing talin proteins with mutated F1 and F2 domains (located in the head) and
discovered that both mutants were unable to form FAs. This means that the F1 and F2
subdomains are essential in focal adhesion assembly and the head domain generally is
necessary. Our cells were expressing the head domain so this explains why the FAs
were assembled quite quickly. Another reason for the quick assembly could be provided
by the fact that minitalins are smaller than the full length talin and they might not be
that flexible to form the autoinhibited conformation. That leads to the assumption that
the active head domain is greatly present during the initiation of the adhesions.
In addition, we could have identified the focal adhesion sites and how fast they
assemble and again disassemble. Berginski et al. (2011) have developed a method to
analyze the FA sites and how they change during time. They used 3T3 fibroblasts with
Total Internal Reflection Fluorescence Microscopy (TIRF). They measured size, shape,
intensity, and position of living cells. With this method it is also possible to focus on the
regulation of FAs and to see how the different talin1 constructs affect the migration and
spreading more precisely on the macromolecular level. It is hard to say if the individual
cell tracking could be more reasonable in this case but it would be interesting to do.
5.5 Cell culture substrates used in this study
This study, as with most in vitro research, was performed in two-dimensional (2D)
substrate although the natural environment of cells is a three-dimensional (3D)
substrate. Because of this, it would be favorable to set the equipment as in a 3D model
but the techniques will need a little more improvement before this can be done.
The majority of the research was done on one type of substrate; fibronectin coated
coverslips. The only exception was the optimization, which was done in a 6-well plate
without any coating. The optimization could have also been done to a coated plate,
although the point was to optimize the method and make it repeatable so coating was
not necessary at this point. Cells are known to form adhesions differently depending on
the substrate (coating and ECM material) and therefore other substrates such as
collagen-I and vitronectin could have also been used. In addition to the ECM materials
in use, there are several other parameters available for modulation. For example, Engler
56
et al. (2004) showed in their publication how the stiffness of the substrate influences the
spreading of myoblasts. Although the plates and the coating used differ significantly
from the physiological environment, the results still give a good assumption from the
cell behavior in vivo.
57
6. CONCLUSIONS
The purpose of this study was to optimize a migration assay and utilize it to study
migration between WT MEF, Tal1-/- and Vin-/- cells. In the second part of the research
the assay was used on Tal1-/- cells transfected with five different talin1 constructs. The
third part comprised of a cell spreading assay with the same transfected Tal1 -/- cells. It
was essential to set up a protocol for the migration in order to get reliable and repeatable
results.
The study presents how focal adhesions can be researched regarding cell migration and
spreading. It also gives a good example on of how various proteins and protein parts
affect forming adhesion sites in ever-changing situations in the cells. Cell migration was
studied with a method called scratch assay and it was performed with talin1-null cells
transfected with five different talin1 constructs. The same cell constructs were also used
for the spreading assay experiments. Most of the parts in this study required
optimization which therefore enables us to use the proven and suitable conditions in the
future experiments as well. Some issues still require more optimization but the protocols
improved here provide a good stepping stone for future studies. At the same time, many
questions arose during the research, such as the migration results of Vin-/- cells,
especially why they migrated slower than their wild type counterparts.
The results we obtained from this study give excellent future prospective opportunities
for the studies focusing on focal adhesions, especially in the fields of cell migration and
spreading. Due to the methodology represented, the study enables the use of mutated
focal adhesion proteins while it offers the study of locating adhesion proteins in the
adhesion sites.
58
REFERENCES
Albiges-Rizo, C., Frachet, P. & Block, M.R. 1995, "Down regulation of talin alters cell
adhesion and the processing of the alpha 5 beta 1 integrin", Journal of cell science,
vol. 108, no. 10, pp. 3317-3329
Anthis, N.J. & Campbell, I.D. 2011, "The tail of integrin activation", Trends in
biochemical sciences, vol. 36, no. 4, pp. 191-198.
Anthis, N.J., Wegener, K.L., Critchley, D.R. & Campbell, I.D. 2010, "Structural
Diversity in Integrin/Talin Interactions", Structure, vol. 18, no. 12, pp. 1654-1666.
Anthis, N.J., Wegener, K.L., Ye, F., Kim, C., Goult, B.T., Lowe, E.D., Vakonakis, I.,
Bate, N., Critchley, D.R., Ginsberg, M.H. & Campbell, I.D. 2009, "The structure of
an integrin/talin complex reveals the basis of inside‐out signal transduction", The
EMBO journal, vol. 28, no. 22, pp. 3623-3632.
Berginski, M.E., Vitriol, E.A., Hahn, K.M. & Gomez, S.M. 2011, "High-resolution
quantification of focal adhesion spatiotemporal dynamics in living cells", PLoS
ONE, vol. 6, no. 7, pp. e22025
Berrier, A.L. & Yamada, K.M. 2007, "Cell-matrix adhesion", Journal of cellular
physiology, vol. 213, no. 3, pp. 565-573.
Calderwood, D.A., Yan, B., de Pereda, J.M., Alvarez, B.G., Fujioka, Y., Liddington,
R.C. & Ginsberg, M.H. 2002, "The Phosphotyrosine Binding-like Domain of Talin
Activates Integrins", Journal of Biological Chemistry, vol. 277, no. 24, pp. 2174921758.
Carisey, A. & Ballestrem, C. 2011, "Vinculin, an adapter protein in control of cell
adhesion signalling", European journal of cell biology, vol. 90, no. 2–3, pp. 157163.
Cohen, D.M., Chen, H., Johnson, R.P., Choudhury, B. & Craig, S.W. 2005, "Two
Distinct Head-Tail Interfaces Cooperate to Suppress Activation of Vinculin by
Talin", Journal of Biological Chemistry, vol. 280, no. 17, pp. 17109-17117.
Cohen, D.M., Kutscher, B., Chen, H., Murphy, D.B. & Craig, S.W. 2006, "A
Conformational Switch in Vinculin Drives Formation and Dynamics of a TalinVinculin Complex at Focal Adhesions", Journal of Biological Chemistry, vol. 281,
no. 23, pp. 16006-16015.
Critchley, D.R. 2000, "Focal adhesions – the cytoskeletal connection", Current opinion
in cell biology, vol. 12, no. 1, pp. 133-139.
Critchley, D.R. 2005, "Genetic, biochemical and structural approaches to talin
function", Biochemical Society transactions, vol. 33, no. Pt 6, pp. 1308-1312.
Critchley, D.R. 2009, "Biochemical and structural properties of the integrin-associated
cytoskeletal protein talin", Annual Review of Biophysics, vol. 38, pp. 235-254
Debrand, E., El Jai, Y., Spence, L., Bate, N., Praekelt, U., Pritchard, C.A., Monkley,
S.J. & Critchley, D.R. "Talin 2 is a large and complex gene encoding multiple
transcripts and protein isoforms", Blackwell Publishing Ltd.
Elliott, P.R., Goult, B.T., Kopp, P.M., Bate, N., Grossmann, J.G., Roberts, G.C.K.,
Critchley, D.R. & Barsukov, I.L. 2010, "The Structure of the Talin Head Reveals a
Novel Extended Conformation of the FERM Domain", Structure, vol. 18, no. 10,
pp. 1289-1299.
Engler, A., Bacakova, L., Newman, C., Hategan, A., Griffin, M. & Discher, D. 2004,
"Substrate compliance versus ligand density in cell on gel responses", Biophysical
journal, vol. 86, no. 1, pp. 617–628
59
Frantz, C., Stewart, K.M. & Weaver, V.M. 2010, "The extracellular matrix at a glance",
Journal of Cell Science, vol. 123, pt. 24, pp. 4195-4200
arc a-Alvarez, B., de Pereda, J.M., Calderwood, D.A., Ulmer, T.S., Critchley, D.,
Campbell, I.D., Ginsberg, M.H. & Liddington, R.C. 2003, "Structural Determinants
of Integrin Recognition by Talin", Molecular cell, vol. 11, no. 1, pp. 49-58.
Geiger, B. & Yamada, K.M. 2011, "Molecular Architecture and Function of Matrix
Adhesions", Cold Spring Harbor Perspectives in Biology, vol. 3, no. 5.
Gingras, A.R., Ziegler, W.H., Bobkov, A.A., Joyce, M.G., Fasci, D., Himmel, M.,
Rothemund, S., Ritter, A., Grossmann, G.J., Patel, B., Bate, N., Goult, B.T.,
Emsley, J., Barsukov, I.L., Roberts, G.C., LIddington, R.C., Ginsberg, M.H. &
Critchley, D.R. 2009, "Structural determinants of integrin binding to the talin rod",
Journal of Biological Chemistry, vol. 284, no. 13, pp. 8866-8876
Gingras, A.R., Bate, N., Goult, B.T., Hazelwood, L., Canestrelli, I., Grossmann, J.G.,
Liu, H., Putz, N.S.M., Roberts, G.C.K., Volkmann, N., Hanein, D., Barsukov, I.L.
& Critchley, D.R. 2007, "The structure of the C‐terminal actin‐binding domain of
talin", The EMBO journal, vol. 27, no. 2, pp. 458-469.
Golji, J. & Mofrad, M.R.K. 2010, "A Molecular Dynamics Investigation of Vinculin
Activation", Biophysical journal, vol. 99, no. 4, pp. 1073-1081.
Goult, B.T., Bate, N., Anthis, N.J., Wegener, K.L., Gingras, A.R., Patel, B., Barsukov,
I.L., Campbell, I.D., Roberts, G.C.K. & Critchley, D.R. 2009, "The Structure of an
Interdomain Complex That Regulates Talin Activity", Journal of Biological
Chemistry, vol. 284, no. 22, pp. 15097-15106.
Goult, B.T., Gingras, A.R., Bate, N., Barsukov, I.L., Critchley, D.R. & Roberts, G.C.K.
2010, "The domain structure of talin: Residues 1815–1973 form a five-helix bundle
containing a cryptic vinculin-binding site", FEBS letters, vol. 584, no. 11, pp.
2237-2241.
Goult, B.T., Xu, X., Gingras, A.R., Swift, M., Patel, B., Bate, N., Kopp, P.M.,
Barsukov, I.L., Critchley, D.R., Volkmann, N. & Hanein, D. 2013, "Structural
studies on full-length talin1 reveal a compact auto-inhibited dimer: Implications for
talin activation", Journal of structural biology, vol. 184, no. 1, pp. 21-32.
Humphries, J.D., Wang, P., Streuli, C., Geiger, B., Humphries, M.J., Ballestrem, C.
2007, " Vinculin controls focal adhesion formation by direct interactions with talin
and actin ", The Journal of Cell Biology, vol. 179, no. 5, pp. 1043-1057
Huttenlocher, A. & Horwitz, A.R. 2011, "Integrins in Cell Migration", Cold Spring
Harbor Perspectives in Biology, vol. 3, no. 9.
Hytönen, V.P. & Vogel, V. 2008, "How force might activate talin’s vinculin binding
sites: SMD reveals a structural mechanism", PLoS Computational Biology, vol. 4,
no. 2, e24
Lee, S.E., Kamm, R.D. & Mofrad, M.R.K. 2007, "Force-induced activation of Talin and
its possible role in focal adhesion mechanotransduction", Journal of Biomechanics,
vol. 40, no. 9, pp. 2096-2106.
Lee, S.E., Chunsrivirot, S., Kamm, R.D. & Mofrad, M.R.K. 2008, "Molecular
Dynamics Study of Talin-Vinculin Binding", Biophysical journal, vol. 95, no. 4,
pp. 2027-2036.
Liang, C., Park, A.Y. & Guan, J. 2007, "In vitro scratch assay: a convenient and
inexpensive method for analysis of cell migration in vitro", Nat.Protocols, vol. 2,
no. 2, pp. 329-333.
Lo, S.H. 2006, "Focal adhesions: What's new inside", Developmental biology, vol. 294,
no. 2, pp. 280-291.
60
Lyman, S., Gilmore, A., Burridge, K., Gidwitz, S. & White, G.C. 1997, "Integrinmediated Activation of Focal Adhesion Kinase Is Independent of Focal Adhesion
Formation or Integrin Activation: studies with activated and inhibitory
ß3cytoplasmic domain mutants", Journal of Biological Chemistry, vol. 272, no. 36,
pp. 22538-22547.
Palecek, S.P., Schmidt, C.E., Lauffenburger, D.A. & Horwitz, A.F. 1996, "Integrin
dynamics on the tail region of migrating fibroblasts", Journal of cell science, vol.
109 (Pt 5), no. Pt 5, pp. 941-952.
Petit, V. & Thiery, J. 2000, "Focal adhesions: structure and dynamics", Biology of the
Cell, vol. 92, no. 7, pp. 477-494.
Praekelt, U., Kopp, P.M., Rehm, K., Linder, S., Bate, N., Patel, B., Debrand, E., Manso,
A.M., Ross, R.S., Conti, F., Zhang, M., Harris, R.C., Zent, R., Critchley, D.R. &
Monkley, S.J. 2012, "New isoform-specific monoclonal antibodies reveal different
sub-cellular localisations for talin1 and talin2", European journal of cell biology,
vol. 91, no. 3, pp. 180-191.
Priddle, H., Hemmings., L., Monkley, S., Woods, A., Patel, B., Sutton, D., Dunn, G.A.,
Zicha, D. & Critchley, D.R. 1998, "Disruption of the Talin Gene Compromises
Focal Adhesion Assembly in Undifferentiated but Not Differentiated Embryonic
Stem Cells", The Journal of Cell Biology, vol. 142, no. 4, pp. 1121-1133
Rao, S.S. & Winter, J.O. 2009, "Adhesion molecule-modified biomaterials for neural
tissue engineering", Frontiers in neuroengineering, vol. 2, pp. 6.
Reszka, A.A., Hayashi, Y. & Horwitz, A.F. 1992, "Identification of amino acid
sequences in the integrin beta 1 cytoplasmic domain implicated in cytoskeletal
association", The Journal of Cell Biology, vol. 117, no. 6, pp. 1321-1330.
Roberts, Gordon C.K. & Critchley, David R. 2009, "Structural and biophysical
properties of the integrin-associated cytoskeletal protein talin" Biophysical reviews,
vol. 1, no. 2, pp. 61-69
Sage, E.H. & Bornstein, P. 1991, "Extracellular proteins that modulate cell-matrix
interactions. SPARC, tenascin, and thrombospondin." Journal of Biological
Chemistry, vol. 266, no. 23, pp. 14831-14834.
Tadokoro, S., Shattil, S.J., Eto, K., Tai, V., Liddington, R.C., de Pereda, J.M., Ginsberg,
M.H. & Calderwood, D.A. 2003, "Talin Binding to Integrin ß Tails: A Final
Common Step in Integrin Activation", Science, vol. 302, no. 5642, pp. 103-106.
Wang, P., Ballestrem, C. & Streuli, C.H. 2011, "The C terminus of talin links integrins
to cell cycle progression", The Journal of Cell Biology, vol. 195, no. 3, pp. 499-513
Wegener, K.L., Partridge, A.W., Han, J., Pickford, A.R., Liddington, R.C., Ginsberg,
M.H. & Campbell, I.D. 2007, "Structural Basis of Integrin Activation by Talin",
Cell, vol. 128, no. 1, pp. 171-182.
Wehrle-Haller, B. 2012, "Assembly and disassembly of cell matrix adhesions", Current
opinion in cell biology, vol. 24, no. 5, pp. 569-581.
Wehrle-Haller, B. 2012, "Structure and function of focal adhesions", Current opinion in
cell biology, vol. 24, no.1, pp. 116-124
Zamir, E. & Geiger, B. 2001, "Components of cell-matrix adhesions", Journal of cell
science, vol. 114, no. 20, pp. 3577-3579.
Zhang X., Jiang G., Cai Y., Monkley S.J., Critchley D.R. & Sheetz M.P. 2008, "Talin
depletion reveals independence of initial cell spreading from integrin activation and
traction", Nature Cell Biology, vol. 10, no. 9, pp. 1062-1068
Xu, W., Baribault, H. & Adamson, E.D. 1998, "Vinculin knockout results in heart and
brain defects during embryonic development", Development, vol. 125, no. 2, pp.
327-337.
61